JP2008240569A - Ignition timing control device for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an ignition timing control device wherein an actual ignition timing does not greatly deviate from an adequate ignition timing even when an ignition timing control model is learned disproportionately in a specified operation region. <P>SOLUTION: The ignition timing control device comprises an initial ignition timing control model which is adapted to calculate such an ignition timing that a combustion condition index value (an 8° combustion rate MFB8) approximately corresponds to a target combustion condition index value. The control device learns the initial ignition timing control model to be corrected to reduce a difference between the axial combustion condition index value and the target combustion condition index value. Then, it acquires a total learning frequency CN and a by-operation-region learning frequency Cx, y and determines a final ignition timing SA(k) from a learning frequency (Cx, y/CN) obtained by dividing the by-operation-region learning frequency corresponding to a current operation region by the total learning frequency, and an initial ignition timing SAin and an after-learning ignition timing SAgk found by the initial ignition timing control model and an after-learning ignition timing control model, respectively. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an ignition timing control device for an internal combustion engine, and more particularly to an ignition timing control device for an internal combustion engine that controls the ignition timing using an ignition timing control model.

  Conventionally, a combustion ratio MFB (Mass Fraction Burned) is calculated based on the in-cylinder pressure (pressure in the combustion chamber) detected by the in-cylinder pressure detecting means, and a predetermined crank angle (for example, a crank angle 8 after compression top dead center) An ignition timing control device for an internal combustion engine is known that controls the ignition timing so that the combustion rate MFB at (°) matches the target combustion rate. Thereby, even when there are individual differences among internal combustion engines, an appropriate ignition timing can be set for each engine. Therefore, the combustion efficiency is improved and the output torque of the internal combustion engine can be increased.

  The combustion ratio MFB is a combustion state index value indicating the combustion state of the engine. The combustion ratio MFB is a value substantially equivalent to the ratio of the indicated heat quantity. The ratio of the indicated amount of heat is as follows: “With respect to a single combustion stroke, the combustion chamber has a predetermined timing with respect to the total amount Qtotal of heat generated by all fuel combusted in the combustion chamber and converted into work for the piston. Is defined as the ratio Qsum / Qtotal of the cumulative amount Qsum of heat converted into work for the piston among the heat generated by the fuel combusted in FIG. The combustion ratio MFB is “of the fuel that contributed to work for the piston among the fuel burned in the combustion chamber by a predetermined timing relative to the total amount of fuel that contributed to work for the piston among all the fuel burned in the combustion chamber”. It is defined as “the percentage of the integrated amount”.

One of such ignition timing control devices acquires an actual combustion ratio at a predetermined crank angle, and advances or retards the ignition timing by a predetermined amount according to the difference between the acquired combustion ratio and the target combustion ratio. (For example, refer to Patent Document 1).
JP-A-9-317522

  By the way, an ignition timing control model is used in place of the method of determining the ignition timing using a map (look-up table) with operating state quantities (for example, engine load and engine speed) representing the operating state of the internal combustion engine as arguments. A technique for determining the ignition timing using the engine has been developed. The ignition timing control model is represented by a function (formula) having the operating state quantity of the internal combustion engine as a variable. This ignition timing control model is preliminarily adapted so that the combustion ratio at a predetermined crank angle matches the target combustion ratio for various operating state quantities. More specifically, the function representing the ignition timing control model is generally a function having a plurality of coefficients, and these coefficients are adapted as described above. In this specification, the ignition timing control model represented by a function (function having a coefficient and the like) adapted in advance is referred to as an “initial ignition timing control model” for convenience.

  On the other hand, there is an airframe difference (individual difference) between individual internal combustion engines. Furthermore, with the use of an internal combustion engine, the characteristics of the engine change over time. Therefore, the optimal ignition timing cannot always be obtained by the initial ignition timing control model. Therefore, a function representing the ignition timing control model is learned by learning so that the actual combustion ratio at a predetermined crank angle matches the target combustion ratio (so that the actual combustion state index value matches the target combustion state index value). It is conceivable to execute learning control to be corrected. That is, in the learning control, the coefficient is sequentially corrected by a known error reduction method such as a sequential least square method so that the difference between the actual combustion ratio and the target combustion ratio at a predetermined crank angle becomes zero. In this specification, the ignition timing control model represented by the function learned in this way (function having a modified coefficient or the like) is referred to as a “post-learning ignition timing control model” for convenience.

  However, as a result of further examination, the inventor has calculated ignition based on a post-learning ignition timing control model in an operation region where learning is insufficient when learning is biased in a specific operation region (learning region). It was found that problems such as knocking occurred and the combustion state became unstable and torque fluctuations occurred due to inappropriate timing.

  Hereinafter, this point will be described in detail. FIG. 13 shows the ignition timing calculated by the ignition timing control model when learning is performed without bias for all the operation regions. That is, the ignition timing obtained by the ignition timing control model shown in FIG. 13 is a substantially appropriate value. On the other hand, FIG. 14 shows the ignition timing calculated by the ignition timing control model after learning is performed only in a specific operation region (for example, a region of medium and low load and medium rotation). From a comparison between FIG. 13 and FIG. 14, if learning is performed biased in the specific operation region, the ignition timing in the specific operation region becomes a value close to an appropriate value, but the ignition timing in other regions is in the specific operation region. It is understood that it is different from the appropriate value under the influence of learning.

  On the other hand, FIG. 15 is a graph showing the number of learning (learning frequency) for each driving region when the vehicle is driven to a certain point. From FIG. 15, it is understood that learning is often performed intensively in a specific driving region.

  From the above, one of the objects of the present invention is that even when learning of the ignition timing control model is intensively performed in a specific operation region, the actual ignition timing may deviate greatly from the appropriate ignition timing. There is no internal combustion engine ignition timing control device.

  An ignition timing control device for an internal combustion engine according to the present invention that achieves the above object includes an initial ignition timing control model. The initial ignition timing control model is a model that is a prototype of the ignition timing control model to be learned. These ignition timing control models are represented by a function for calculating the ignition timing of the engine with the operating state quantity representing the operating state of the engine as a variable. In the initial ignition timing control model, when ignition is performed at the ignition timing calculated by the initial ignition timing control model, a combustion state index value indicating a combustion state of the engine is a target combustion state index value indicating a predetermined target combustion state. Pre-adapted to match.

  The ignition timing control device further includes an operating state quantity acquisition unit, an initial ignition timing calculation unit, a combustion state index value acquisition unit, a model learning unit, and a post-learning ignition timing calculation unit.

The operation state quantity acquisition means acquires the actual operation state quantity as the actual operation state quantity. The operating state quantity is, for example, an engine load and an engine speed.
The initial ignition timing calculation means calculates the initial ignition timing by applying the acquired actual operation state quantity to the initial ignition timing control model.

The combustion state index value acquisition means acquires an actual combustion state index value as an actual combustion state index value. The combustion state index value is, for example, a combustion ratio at a predetermined crank angle, a ratio of indicated heat quantity at a predetermined crank angle, an in-cylinder pressure at a predetermined crank angle, and the like.
The model learning means represents a function representing the initial ignition timing control model so that a difference between the actual combustion state index value acquired when the actual combustion state index value is acquired and the target combustion state index value is reduced. Learning that sequentially corrects (that is, further corrects the ignition timing control model obtained at that time) is performed.
The post-learning ignition timing calculation means applies the acquired actual operating state quantity to the ignition timing control model (that is, the post-learning ignition timing control model) learned by the model learning means, thereby calculating the post-learning ignition timing. calculate.

  In addition, the ignition timing control device includes a total number of learning times acquisition unit, a learning number-by-operation region acquisition number acquisition unit, and an ignition timing control unit.

The total learning number acquisition means acquires the total learning number (the total number of learning times by the model learning means) by accumulating the learning times by the model learning means.
The learning count by operation region acquisition means divides the operation region of the engine into a plurality of operation regions in advance, and indicates the number of times of learning by the model learning device by the actual operation state quantity when the learning is performed. The number of learning for each driving region is acquired by integrating the driving state corresponding to one driving region (learning driving region) of the plurality of driving regions to which the driving state belongs. In other words, the learning number by driving region acquisition unit acquires the learning number by the model learning unit as a learning number by driving region for each predetermined driving region. As a result, when the learning is performed, the number of learning for each driving region corresponding to the learning driving region is increased (incremented).

  The ignition timing control means is “current operation region (one operation region of the plurality of operation regions to which the operation state (current operation state of the engine) indicated by the actual operation state quantity at the present time belongs)”. The learning ratio which is a value corresponding to the ratio of the number of learnings for each driving region acquired corresponding to “current driving region)” and the total number of learnings is acquired. For example, the learning ratio is the ratio of the number of learnings for each driving region corresponding to the current driving region to the total number of learnings (learning ratio = number of learnings for each driving region in the current driving region / total number of learnings). Further, the ignition timing control means obtains a final ignition timing reflecting the initial ignition timing and the post-learning ignition timing in accordance with the learning ratio, and executes ignition at the obtained final ignition timing.

  The learning ratio represents the accuracy of the after-learning ignition timing model in the current operation region. That is, the learning ratio represents the reliability of the post-learning ignition timing. On the other hand, since the initial ignition timing is an ignition timing calculated by an initial ignition timing control model adapted in advance, the reliability thereof is somewhat high.

  Therefore, the ignition timing control means obtains a final ignition timing reflecting the initial ignition timing and the post-learning ignition timing according to the learning ratio, and performs ignition at the final ignition timing. As a result, even if the learning timing in the current operation region is deviated from the appropriate value as a result of intensive learning in the operation region different from the current operation region, the final ignition timing is greatly increased from the appropriate value. Deviation can be avoided. Furthermore, if sufficient learning is performed in the current operation region, a highly reliable post-learning ignition timing can be set as the final ignition timing or greatly reflected in the final ignition timing. It becomes possible to execute actual ignition at the timing.

  In this case, the combustion state index value is an actual combustion ratio when the crank angle of the engine is a predetermined crank angle, and the target combustion state index value is the crank angle of the engine having the predetermined crank angle. It is preferable that the target value of the combustion rate at that time.

  According to this, even when there is an individual difference between engines, an appropriate ignition timing can be set for each engine. Therefore, the combustion efficiency is improved and the output torque of the internal combustion engine can be increased.

Further, the ignition timing determining means includes
The post-learning ignition timing is determined as the final ignition timing when the learning ratio for the current operation region (current operation region) is larger than a predetermined ratio, and the initial ignition timing when the learning ratio is smaller than the predetermined ratio. May be configured to determine the final ignition timing.

  According to this, when the learning ratio is larger than the predetermined ratio, that is, when the post-learning ignition timing is close to an appropriate value, actual ignition is performed at the post-learning ignition timing. As a result, the actual combustion state index value becomes a value close to the target combustion state index value, so that the combustion state can be made closer to the target combustion state. Further, when the learning ratio is smaller than the predetermined ratio, that is, when there is a high possibility that the post-learning ignition timing is greatly deviated from the appropriate value, actual ignition is performed at the initial ignition timing. As a result, it is possible to avoid the ignition timing from greatly deviating from an appropriate timing, so that combustion can be stabilized.

  On the other hand, the ignition timing control means may be configured to obtain the final ignition timing by weighting the post-learning ignition timing and the initial ignition timing according to a learning ratio for the current operation region. Good. In this case, the ignition timing control means obtains the final ignition timing so that the weight of the post-learning ignition timing increases and the weight of the initial ignition timing decreases as the learning ratio for the current operating region increases.

  According to this, the higher the reliability of the ignition timing after learning (the closer the ignition timing after learning is to an appropriate value), the more the ignition timing after learning is reflected in the final ignition timing. In other words, the lower the reliability of the after-learning ignition timing, the more the initial ignition timing is reflected in the final ignition timing. As a result, actual ignition can be executed at an ignition timing closer to an appropriate value.

  The ignition timing control device for another internal combustion engine according to the present invention is similar to the ignition timing control device described above, in that the initial ignition timing control model, the operating state quantity acquisition means, the initial ignition timing calculation means, and the combustion state index value acquisition means. And a model learning means and a post-learning ignition timing calculation means.

And the ignition timing control device is
Load change speed acquisition means for acquiring a change speed of a load of the engine (for example, throttle valve opening, accelerator pedal operation amount, intake air amount in one intake stroke of the engine, etc.);
When the obtained load change speed (change speed magnitude) is smaller than a predetermined load change speed threshold, the post-learning ignition timing is determined as the final ignition timing, and the acquired load change speed is the same load. An ignition timing control means for determining the initial ignition timing as a final ignition timing when greater than a change speed threshold, and executing ignition at the determined final ignition timing;
It has.

  The learning is control for correcting the ignition timing control model by feedback control while leaving past information. Therefore, learning cannot correct the ignition timing control model with sufficient responsiveness during transient operation such as rapid acceleration when the engine load changes rapidly. Therefore, as described above, when the load change rate is larger than the predetermined load change rate threshold, that is, when the correction of the ignition timing control model by learning is not in time, the initial ignition timing is substituted for the post-learning ignition timing. If ignition is performed in this manner, it is possible to avoid the ignition timing from greatly deviating from the appropriate value.

  Still another ignition timing control device for an internal combustion engine according to the present invention is similar to the ignition timing control device described above. An initial ignition timing control model, an operating state quantity acquisition means, an initial ignition timing calculation means, and a combustion state index value acquisition. Means, model learning means, and post-learning ignition timing calculation means.

  When the difference between the learned ignition timing and the initial ignition timing calculation (the magnitude of the difference, the absolute value of the difference) is smaller than a predetermined threshold, the ignition timing control device uses the learned ignition timing as the final ignition timing. When the difference is larger than the predetermined threshold value, there is provided an ignition timing control means for determining a limited ignition timing determined based on the initial ignition timing as a final ignition timing and executing ignition at the determined final ignition timing. .

  The limited ignition timing is set to an ignition timing obtained by advancing the initial ignition timing by a first predetermined amount when the post-learning ignition timing is on the advance side with respect to the initial ignition timing. If it is on the retard side, the initial ignition timing may be set to an ignition timing retarded by a second predetermined amount. The first predetermined amount may be equal to or different from the second predetermined amount. In order to suppress knocking due to excessive advance of the ignition timing, the first predetermined amount can be made smaller than the second predetermined amount. Further, the predetermined threshold value may be constant, or may be a value that changes according to the amount of operating state of the engine. In particular, in order to avoid occurrence of knocking on the high load side, the predetermined threshold value may be set so as to decrease as the engine load increases.

  As described above, the initial ignition timing control model is adapted in advance. Therefore, the initial ignition timing is not so different from the proper ignition timing. Therefore, when the difference between the post-learning ignition timing and the initial ignition timing calculation is greater than a predetermined threshold, the ignition timing control model is affected by the learning in the other operating regions, so that the post-learning ignition timings in the current operating region Can be considered to deviate significantly from the appropriate value. Therefore, the ignition timing control means determines the limited ignition timing determined based on the initial ignition timing as the final ignition timing when the difference between the post-learning ignition timing and the initial ignition timing calculation is greater than a predetermined threshold, When the difference between the post-learning ignition timing and the initial ignition timing calculation is smaller than a predetermined threshold, the post-learning ignition timing is determined as the final ignition timing, and ignition is performed at the final ignition timing determined as such. As a result, it is possible to avoid the ignition timing from greatly deviating from the appropriate value.

Hereinafter, an ignition timing control device for an internal combustion engine according to each embodiment of the present invention will be described with reference to the drawings.
<First Embodiment>
(Constitution)
FIG. 1 shows a spark ignition type multi-cylinder (four-cylinder) 4 of a piston reciprocating type which is an ignition timing control device (hereinafter also referred to as “first control device”) according to a first embodiment of the present invention. 1 shows a schematic configuration of a system applied to a cycle internal combustion engine 10. Although FIG. 1 shows only a cross section of a specific cylinder, other cylinders have the same configuration.

  The internal combustion engine 10 includes a cylinder block portion 20 including a cylinder block, a cylinder block lower case, an oil pan, and the like, a cylinder head portion 30 fixed on the cylinder block portion 20, and a gasoline mixture to the cylinder block portion 20. An intake system 40 for supplying and an exhaust system 50 for releasing exhaust gas from the cylinder block 20 to the outside are included.

  The cylinder block unit 20 includes a cylinder 21, a piston 22, a connecting rod 23, and a crankshaft 24. The piston 22 reciprocates in the cylinder 21, and the reciprocating motion of the piston 22 is transmitted to the crankshaft 24 through the connecting rod 23, whereby the crankshaft 24 rotates. The heads of the cylinder 21 and the piston 22 form a combustion chamber 25 together with the cylinder head portion 30.

  The cylinder head unit 30 includes an intake port 31 that communicates with the combustion chamber 25, an intake valve 32 that opens and closes the intake port 31, an intake valve control device 33 that opens and closes the intake valve 32, an exhaust port 34 that communicates with the combustion chamber 25, an exhaust An exhaust valve 35 that opens and closes the port 34, an exhaust camshaft 36 that drives the exhaust valve 35, an ignition plug 37, an igniter 38 that includes an ignition coil that generates a high voltage applied to the ignition plug 37, and fuel are injected into the intake port 31. An injector (fuel injection means) 39 is provided.

  The intake valve control device 33 has a known configuration that adjusts and controls the relative rotation angle (phase angle) between an intake cam shaft and an intake cam (not shown) by hydraulic pressure, and opens the intake valve 32 (intake valve). The valve opening time) can be changed. In this example, the valve opening period (the valve opening crank angle width) of the intake valve is constant. Therefore, when the intake valve opening timing is advanced or retarded by a predetermined angle, the intake valve closing timing is also advanced or retarded by the predetermined angle. Further, the opening timing and closing timing of the exhaust valve 35 are constant. Accordingly, the overlap period changes as the intake valve opening timing is changed by the intake valve control device 33.

  The intake system 40 includes an intake manifold 41 including an intake manifold that communicates with the intake port 31 and forms an intake passage together with the intake port 31, an air filter 42 provided at an end of the intake pipe 41, and an intake pipe 41. A throttle valve 43 having a variable opening cross-sectional area of the passage and a throttle valve actuator 43a including a DC motor constituting throttle valve driving means are provided.

  The exhaust system 50 includes an exhaust manifold 51 communicating with the exhaust port 34, an exhaust pipe (exhaust pipe) 52 connected to the exhaust manifold 51, an upstream three-way catalyst 53, and a downstream three-way catalyst 54. The upstream three-way catalyst 53 is disposed in the exhaust pipe 52. The downstream three-way catalyst 54 is disposed in the exhaust pipe 52 downstream of the upstream three-way catalyst 53. The exhaust port 34, the exhaust manifold 51, and the exhaust pipe 52 constitute an exhaust passage.

  On the other hand, this system includes a hot-wire air flow meter 61, a throttle position sensor 62, a cam position sensor 63, a crank position sensor 64, an in-cylinder pressure sensor 65 provided in each cylinder, a coolant temperature sensor 66, and an upstream of the first catalyst 53. An air-fuel ratio sensor 67 disposed in the exhaust passage, an air-fuel ratio sensor 68 and an accelerator opening sensor 69 disposed in the exhaust passage downstream of the first catalyst 53 and upstream of the second catalyst 54 are provided.

  The hot-wire air flow meter 61 detects the mass flow rate per unit time of the intake air flowing through the intake pipe 41 and outputs a signal representing the mass flow rate Ga. The throttle position sensor 62 detects the opening of the throttle valve 43 and outputs a signal representing the throttle valve opening TA. The cam position sensor 63 outputs one pulse every time the intake camshaft rotates 90 degrees, 90 degrees, and 180 degrees from a predetermined angle. This signal is also referred to as a G2 signal. The crank position sensor 64 outputs a pulse every time the crankshaft 24 rotates 10 degrees. The pulse output from the crank position sensor 64 is converted into a signal representing the engine speed NE. Further, the crank angle of the engine 10 is obtained based on signals from the cam position sensor 63 and the crank position sensor 64. The in-cylinder pressure sensor 65 detects the pressure in the combustion chamber 25 and outputs a signal representing the in-cylinder pressure Pc.

  The upstream air-fuel ratio sensor 67 and the downstream air-fuel ratio sensor 68 detect the upstream and downstream air-fuel ratios of the catalyst 53 and output signals representing the upstream and downstream air-fuel ratios, respectively. The accelerator opening sensor 69 detects the operation amount of the accelerator pedal 81 operated by the driver, and outputs a signal representing the operation amount Accp of the accelerator pedal 81.

  The electric control device 70 includes a CPU 71 connected by a bus, a routine (program) executed by the CPU 71, a ROM 72 pre-stored with tables (look-up tables, maps), constants, and the like, and the CPU 71 temporarily stores data as necessary. The microcomputer 73 includes a RAM 73 that stores data, a backup RAM 74 that stores data while the power is on, and retains the stored data even when the power is shut off, and an interface 75 that includes an AD converter. The interface 75 is connected to the sensors 61 to 69 and supplies signals from the sensors 61 to 69 to the CPU 71. The interface 75 sends a drive signal to the intake valve control device 33, the injector 39, and the throttle valve actuator 43a in accordance with an instruction from the CPU 71, and sends an ignition signal to the igniter 38.

(Outline of control)
Next, an outline of ignition timing control performed by the ignition timing control device (first control device) of the internal combustion engine 10 configured as described above will be described.

<Estimation (acquisition) of combustion ratio MFB>
The combustion ratio MFB defined as described above is estimated (acquired) as a value representing the ratio Qsum / Qtotal of the indicated heat quantity defined as described above. The combustion ratio MFB and the ratio of the indicated heat quantity are both combustion state index values indicating the combustion state of the engine 10. Details of the method for obtaining the combustion ratio MFB from the in-cylinder pressure Pc detected by the in-cylinder pressure sensor 65 are disclosed in, for example, Japanese Patent Application Laid-Open No. 2006-144645, and the outline thereof will be described below.

  In this example, the combustion ratio MFB is obtained corresponding to the crank angle θ representing a predetermined timing. The combustion ratio MFB at the crank angle θ is expressed as MFBθ. This crank angle θ is 0 at the compression top dead center, takes a negative value that increases in absolute value as it advances from the compression top dead center toward the compression top dead center, and from the compression top dead center to the compression top dead center. It is defined to take a positive value, the absolute value of which increases as the angle is retarded later. For example, θ = −θ1 ° (θ1> 0) indicates that the crank angle is BTDCθ1. θ = θ2 ° (θ2> 0) indicates that the crank angle is ATDCθ2.

The combustion ratio MFBθ at the crank angle θ is estimated by the following equation (1). In the equation (1), the crank angle θs (θs <0) is set so that both the intake valve 32 and the exhaust valve 35 are closed in the compression stroke toward the target combustion stroke (expansion stroke) and is higher than the ignition timing. It is a time when the angle is sufficiently advanced (for example, θs = −60 °, that is, BTDC 60 °). The crank angle θe (θe> 0) is a predetermined timing later than the latest timing at which combustion in the target combustion stroke is substantially ended and a timing advanced from the exhaust valve opening timing (for example, θe = 60). °, ie ATDC 60 °).

This equation (1) is based on the knowledge that the change pattern of the accumulated amount Q of heat contributed to the work for the piston among the generated heat is substantially coincident with the change pattern of Pc (θ) V (θ) ^κ . Pc (θ) is the in-cylinder pressure at the crank angle θ, V (θ) is the volume of the combustion chamber 25 at the crank angle θ, and κ is the specific heat ratio (for example, 1.32) of the mixed gas. The denominator of equation (1) is a value corresponding to 100% of MFB.

<Basic contents of ignition timing control>
First, the basic contents of ignition timing control by the first control device will be described. The details of the first control device learning (learning control) and the final ignition timing (final ignition timing) determination method will be described later.

  FIG. 2 is a graph showing the relationship between the ignition timing SA, the 8 ° combustion ratio MFB8, and the generated torque TRQ of the engine 10. The 8 ° combustion rate MFB8 is the combustion rate MFB when the crank angle θ is 8 ° (= ATDC 8 °) after compression top dead center. As is apparent from FIG. 2, the 8 ° combustion ratio MFB8 at which the generated torque TRQ is maximum is about 60% (see region A in FIG. 2). Accordingly, the first control device controls the ignition timing so that the 8 ° combustion ratio MFB8 becomes a predetermined target value MFB8tgt (for example, a value close to 60%).

  More specifically, the first control device has each function achievement unit shown in FIG. 3 which is a functional block diagram. Hereinafter, the function of each block will be described in order. In this specification, the symbol (k) given after a variable indicates that the variable is a variable for a combustion stroke that occurs next in a specific cylinder (that is, the current combustion stroke). Therefore, the variable to which (k-1) is assigned is a variable for the previous combustion stroke (the combustion stroke immediately before the end of the combustion, that is, the combustion stroke one time before (720 ° crank angle)) in the specific cylinder. It is. Further, the ignition timing SA (SA> 0) means that ignition is performed at BTDC SA ° (crank angle SA ° before compression top dead center).

  The operation state amount acquisition unit A1 acquires an amount (operation state amount) representing the operation state of the engine. In this example, the operating state quantities are the load KL and the engine speed NE of the engine 10. The load KL is a value (filling rate) proportional to the in-cylinder air weight sucked, and is obtained from the mass flow rate Ga detected by the air flow meter 61 and the engine speed NE. The load KL may be acquired by a well-known air model describing the behavior of air. The operating state quantity acquisition unit A1 may be configured to acquire the operation amount Accp of the accelerator pedal 81, the throttle valve opening degree TA, and the like as the load of the engine 10 instead of the load (filling rate) KL.

  The target value setting unit A2 outputs “target combustion ratio MFB8tgt” which is a target value for the 8 ° combustion ratio MFB8. The target combustion ratio MFB8tgt is a constant value (for example, 60%) in this example. The target value setting unit A2 may be configured to input an operation state quantity and change the target combustion ratio MFB8tgt according to the operation state quantity. The target combustion ratio MFB8tgt is set to such a value that the combustion efficiency of the engine is good, the emission amount of HC, CO, etc. is low, and torque fluctuation due to knocking or the like does not occur.

The actual combustion ratio MFB8 calculating unit A3 calculates the combustion ratio MFB8 (k−1) at ATDC 8 ° during the previous combustion stroke according to the above equation (1). That is, the actual combustion ratio MFB8 calculating means A3 calculates the 8 ° combustion ratio MFB8 (k−1) based on the following equation (2). Each value on the right side in equation (2) is a value obtained during the previous combustion stroke. Note that −60 ° means a crank angle of 60 ° before compression top dead center.

  The ignition timing deviation amount calculation unit A4 calculates the difference ΔM between the target combustion rate MFB8tgt acquired from the target value setting unit A2 and the actual 8 ° combustion rate MFB8 (k−1) acquired from the actual combustion rate MFB8 calculation unit A3. Based on the above, the amount by which the current ignition timing SA (k) should be advanced with respect to the previous ignition timing SA (k−1) (the crank angle to be advanced, ie, the ignition timing correction amount or the ignition timing deviation amount) ΔSA (K-1) is calculated. Specifically, the ignition timing deviation amount calculation unit A4 includes a table (or function) that predefines the relationship between the difference ΔM and the ignition timing deviation amount ΔSA. The ignition timing deviation amount calculation unit A4 calculates the actual ignition timing deviation amount ΔSA (k−1) by applying the actually obtained difference ΔM to the table.

  The delay unit A5 holds the operation state amount (NE (k), KL (k)) for the current combustion stroke acquired by the operation state amount acquisition unit A1, and is used to determine the ignition timing for the previous combustion stroke. The operation state quantity (NE (k-1), KL (k-1)) for the previous combustion stroke is output.

  The model learning unit A6 calculates the ignition timing deviation amount ΔSA (k−1) obtained from the ignition timing deviation amount calculation unit A4 and the previous operating state amount (NE (k−1), KL (k) obtained from the delay unit A5. -1)), the ignition timing control model is learned. More specifically, the model learning unit A6 sets the coefficients θ0 to θ3 of a function representing an ignition timing control model, which will be described later, so that the ignition timing deviation amount ΔSA (k−1) becomes a minimum value. It calculates based on a recursive least squares (RLS).

The ignition timing control unit A7 has an ignition timing control model represented by the function of the following equation (3).

  As described above, the coefficients θ0 to θ3 in the equation (3) are corrected (learned) by the model learning unit A6. The coefficients θ0 to θ3 are “8 ° combustion ratio MFB8 is a target combustion ratio for various engine speeds NE and engine loads KL before the model learning unit A6 corrects them (that is, initial state). It is preliminarily adapted by experiment or the like so as to obtain an ignition timing SA (k) that substantially matches MFB8tgt. The ignition timing control unit A7 holds the function represented by the equation (3) using the coefficient (initial coefficient) before correction as a function representing the “initial ignition timing control model”. Further, the ignition timing control unit A7 holds the function represented by the equation (3) using the modified coefficient (the post-learning coefficient) as a function representing the “post-learning ignition timing control model”. . The initial coefficient and the post-learning coefficient are stored in the backup RAM 74.

  Then, the ignition timing control unit A7 of the first control device calculates the initial ignition timing SAin by applying the operation state quantities (NE (k), KL (k)) for the current combustion stroke to the initial ignition timing control model. At the same time, the after-learning ignition timing SAgk is calculated by applying the operating state quantities (NE (k), KL (k)) for the current combustion stroke to the after-learning ignition timing control model. Further, as will be described later, the ignition timing control unit A7 determines a final ignition timing (final ignition timing) SA (k) using the initial ignition timing SAin and the post-learning ignition timing SAgk, and the final ignition timing. An ignition signal (ignition instruction signal) is sent to the igniter 38 so that ignition is performed at.

<Details of ignition timing control>
Next, details of ignition timing control by the first control device will be described. The first control device acquires the total number of times of learning by the model learning unit A6 (the number of times the coefficients θ0 to θ3 have been corrected by learning) as the total number of learning times. Further, the first control device acquires the number of learning by the model learning unit A6 as the number of learning for each driving region when the learning is performed.

  Then, the first control device learns by operation region corresponding to the operation region (current operation region) to which the current operation state indicated by the current operation state amount (the operation state amount at the time of determining the current ignition timing) belongs. The number of times is read, and a learning ratio that is the ratio of the number of learning times for each driving region to the total number of learning times acquired is obtained. Further, the first control device obtains a final ignition timing reflecting the initial ignition timing and the post-learning ignition timing according to the learning ratio, and executes ignition at the obtained final ignition timing.

  More specifically, the first control device determines the post-learning ignition timing as the final ignition timing when the learning ratio for the current operating region is greater than a predetermined ratio, and the learning ratio is greater than the predetermined ratio. Is smaller, the initial ignition timing is determined as the final ignition timing. Hereinafter, the actual operation of the first control device will be described with reference to a flowchart.

  The CPU 71 of the electric control device 70 executes an in-cylinder pressure acquisition routine (not shown) so that the Pc (−60 °), Pc (8 °), and Pc (60 °) of each cylinder are based on the output of the in-cylinder pressure sensor 65. And stored in the RAM 73 for each cylinder. Further, the CPU 71 performs the ignition timing control routine shown in the flowchart of FIG. 4 with a predetermined crank angle immediately before the start of the combustion stroke when the crank angle of the specific cylinder (for example, crank angle 90 ° before compression top dead center of the specific cylinder = BTDC). It is repeatedly executed every time it matches 90 °. The CPU 71 executes the same routine for other cylinders at the same timing.

  When the crank angle of the specific cylinder coincides with the predetermined crank angle, the CPU 71 starts the processing of the routine of FIG. 4 from step 400, and performs the following processing from step 405 to step 460.

  Step 405: The CPU 71 calculates the actual 8 ° combustion ratio MFB8 (k−1) in the previous combustion stroke of the specific cylinder based on the above equation (2). At this time, the CPU 71 applies Pc (−60 °), Pc (8 °), and Pc (60 °) of the specific cylinder stored in the RAM 73 to the equation (2). The CPU 71 assigns values stored in the ROM 72 in advance to the combustion chamber volumes V (−60 °), V (8 °), and V (60 °) of the equation (2).

  Step 410: The CPU 71 obtains the engine speed NE (k−1) and the load KL (k−1) (used when determining the previous ignition timing of the specific cylinder) for the previous combustion stroke of the specific cylinder. . At the same time, the CPU 71 obtains the engine speed NE (k) and the load KL (k) for the current combustion stroke of the specific cylinder (used when determining the current ignition timing of the specific cylinder). The engine speed NE (k) and the load KL (k) are equal to the engine speed NE and the load KL at the current time, respectively.

Step 415: The CPU 71 determines the difference ΔM between the predetermined target combustion rate MFB8tgt and the actual 8 ° combustion rate MFB8 (k−1) acquired in Step 405 (from the target combustion rate MFB8tgt to 8 ° combustion rate MFB8 (k A value ΔM) obtained by subtracting -1) is obtained.
Step 420: The CPU 71 calculates the ignition timing deviation amount ΔSA (k−1) by applying the difference ΔM to the function f (or look-up table). In this example, the function f is a function that multiplies the difference ΔM by a constant k (or a monotonically increasing function with respect to the difference ΔM).

  Step 425: The CPU 71 executes learning of the ignition timing control model. Specifically, the CPU 71 calculates (determines) the coefficients θ0, θ1, θ2, and θ3 that minimize the ignition timing deviation amount ΔSA (k−1) based on the sequential least square method. At this time, the CPU 71 uses the engine speed NE (k−1) and the load KL (k−1) for calculation. The calculated new coefficients θ0, θ1, θ2, and θ3 are stored in the backup RAM 74 as post-learning coefficients.

  Step 430: The CPU 71 is a learning counter that identifies an operation region (learning operation region) to which the engine speed NE (k-1) and the load KL (k-1) belong, and is specific to the identified operation region. The value of the learning counter for each operation area Cm, n is read. More specifically, as shown in FIG. 5, the operation region (learning region) is divided into a plurality of regions according to the engine speed NE and the load KL. The CPU 71 specifies in which of the plurality of operation areas shown in FIG. 5 the operation state indicated by the engine speed NE (k-1) and the load KL (k-1) exists. Then, the value of the learning area counter Cm, n assigned to the identified operation area is read.

  Step 435: The CPU 71 increases the value of the learning region-specific learning counter Cm, n read in Step 430 by “1”. As a result, the value of the learning region-by-operation region counter Cm, n is the number of times the ignition timing control model is learned in the operation region to which the learning number counter Cm, n is assigned (coefficients θ0 to θ3 are corrected). This is a value representing the number of times, that is, the number of times of learning for each operation region of the ignition timing control model.

Step 440: The CPU 71 increases the value of the total learning number counter CN by “1”. As a result, the value of the total learning number counter CN is a value representing the total number of times that the ignition timing control model has been learned (the total number of times that the coefficients θ0 to θ3 are corrected regardless of the operation region, that is, the total number of learning times). It becomes.
Step 445: The CPU 71 specifies the operation region to which the engine speed NE (k) and the load KL (k) belong, and reads the value of the learning counter for each operation region Cx, y in the specified region. The operating region to which the engine speed NE (k) and load KL (k) (the current operating state of the engine indicated by the current operating state amount) belongs is referred to as a “current operating region”.

  Step 450: The CPU 71 sets the coefficients θ0 to θ3 in the above expression (3) to the coefficients θ0 to θ3 updated (learned) in step 425, and the engine speed NE (k) is set in the expression (3). Then, by substituting the load KL (k), the ignition timing SAgk corresponding to the current ignition timing (ignition timing for the next combustion stroke) is calculated. The ignition timing obtained by the ignition timing control model (post-learning ignition timing control model) in which the coefficients θ0 to θ3 are set to the learned values (post-learning coefficients) in this way is also referred to as “post-learning ignition timing”.

  Step 455: The CPU 71 sets the coefficients θ0 to θ3 in the above expression (3) to initial values (initial coefficients), and substitutes the engine speed NE (k) and the load KL (k) into the expression (3). Thus, the ignition timing SAin corresponding to the current ignition timing is calculated. The ignition timing obtained by the ignition timing control model (initial ignition timing control model) in which the coefficients θ0 to θ3 are returned to the initial values in this way is also referred to as “initial ignition timing”.

  Step 460: The CPU 71 obtains a learning ratio (= Cx, y / CN) by dividing the learning counter for each operation region Cx, y read in step 445 by the total learning number counter CN. Then, the CPU 71 determines whether or not the learning ratio (= Cx, y / CN) is larger than a threshold value (predetermined ratio) Th.

  Now, it is assumed that the learning ratio (= Cx, y / CN) is larger than the threshold Th. That the learning ratio (= Cx, y / CN) is larger than the threshold value Th is relatively sufficient in the current operating region to which the current operating state represented by the engine speed NE (k) and the load KL (k) belongs. This means that the learning is performed or that the number of times of learning in the current operation region is small, but the learning in other operation regions has a small influence on the ignition timing control model in the current operation region. That is, in this case, the post-learning ignition timing SAgk is considered to be a highly reliable value (a value close to the appropriate value). Accordingly, the CPU 71 makes a “Yes” determination at step 460 to proceed to step 465 where the post-learning ignition timing SAgk obtained at step 450 is set to the final ignition timing (ignition timing for the current combustion stroke) SA (k). Set.

  Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) set in step 465 (that is, the post-learning ignition timing SAgk). . Then, the CPU 71 proceeds to step 495 and once ends this routine.

  On the other hand, when the learning ratio (= Cx, y / CN) is smaller than the threshold value Th (when the learning ratio is equal to or less than the threshold value Th), learning in other operation regions has a great influence on the ignition timing control model in the current operation region. Means that In other words, in this case, the initial ignition timing SAin is more reliable than the post-learning ignition timing SAgk. Accordingly, the CPU 71 makes a “No” determination at step 460 to proceed to step 475 to set the initial ignition timing SAin obtained at step 455 to the final ignition timing SA (k).

  Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) set in step 475 (that is, the initial ignition timing SAin). Then, the CPU 71 proceeds to step 495 and once ends this routine.

  As described above, the first control device determines the post-learning ignition timing SAgk as the final ignition timing when the learning ratio (= Cx, y / CN) for the current operation region is larger than the predetermined ratio (Th). SA (k) is determined, and when the learning ratio (= Cx, y / CN) is smaller than a predetermined ratio (Th), the initial ignition timing SAin is determined as the final ignition timing SA (k). Ignition timing control means.

  Therefore, when the learning ratio (= Cx, y / CN) is larger than the predetermined ratio (Th), that is, when the post-learning ignition timing SAgk is close to an appropriate value, the actual ignition is performed at the post-learning ignition timing SAgk. Done. As a result, the actual combustion state index value (8 ° combustion ratio MFB8) is close to the target combustion state index value (target combustion ratio MFB8tgt), so that the combustion state can be made closer to the target combustion state. Thereby, the efficiency of the engine can be kept high.

  Further, when the learning ratio (= Cx, y / CN) is smaller than the predetermined ratio (Th), that is, when it is highly possible that the post-learning ignition timing SAgk is greatly deviated from an appropriate value, the initial ignition timing is set. Actual ignition is performed at SAin. As a result, it is possible to avoid the ignition timing from greatly deviating from an appropriate timing, so that combustion can be stabilized.

Second Embodiment
Next, an ignition timing control device (hereinafter referred to as “second control device”) according to a second embodiment of the present invention will be described. The second controller differs from the first controller only in the method for determining the final ignition timing. Therefore, hereinafter, this difference will be mainly described.

  The CPU 71 of the second control device performs the ignition timing control routine shown in the flowchart of FIG. 6 instead of FIG. 4 every time the crank angle of a specific cylinder matches a predetermined crank angle (for example, BTDC 90 °) immediately before the start of the combustion stroke. Repeatedly executed. In FIG. 6, steps for performing the same processing as the steps shown in FIG. 4 are denoted by the same reference numerals as those assigned to such steps in FIG. 4. A detailed description of these steps is omitted.

  When the crank angle of the specific cylinder coincides with the predetermined crank angle, the CPU 71 starts the processing of the routine of FIG. 6 from step 600 and performs the processing of step 405 to step 455 described above. As a result, the ignition timing control model is corrected by learning based on the previous combustion (step 425), and the learning region-specific learning counter Cm, n corresponding to the learned operation region is increased (step 435). Then, the value of the total learning number counter CN is increased (step 440). In addition, a post-learning ignition timing SAgk and an initial ignition timing SAin are calculated (step 450, step 455).

Next, the CPU 71 proceeds to step 605 to calculate the final ignition timing SA (k) based on the following equation (4).

  That is, in step 605, the CPU 71 increases the weight of the post-learning ignition timing SAgk and increases the initial ignition timing as the learning ratio (= Cx, y / CN) for the current operation region (current operation region) increases. In order to reduce the weight of SAin, the post-learning ignition timing SAgk and the initial ignition timing SAin are weighted according to the learning ratio (= Cx, y / CN). That is, using the post-learning ignition timing SAgk, the initial ignition timing SAin, and the learning ratio (= Cx, y / CN), the weighted average value of the post-learning ignition timing SAgk and the initial ignition timing SAin is obtained, and the weighted average value is used as the final ignition. Set to time SA (k).

  Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) determined in step 605. Then, the CPU 71 proceeds to step 495 and once ends this routine.

  As described above, the second control device weights the post-learning ignition timing SAgk and the initial ignition timing SAin according to the learning ratio (= Cx, y / CN), thereby making the final ignition timing. Ignition timing control means for calculating SA (k) is provided.

  According to this, the higher the reliability of the post-learning ignition timing SAgk (the closer the post-learning ignition timing SAgk is to an appropriate value), the more the post-learning ignition timing SAgk is reflected in the final ignition timing SA (k). In other words, the lower the reliability of the after-learning ignition timing SAgk, the more the initial ignition timing SAin is reflected in the final ignition timing SA (k). As a result, actual ignition can be executed at an ignition timing closer to an appropriate value. Accordingly, the efficiency of the engine can be maintained high, and the ignition timing can be prevented from greatly deviating from the appropriate timing, so that combustion can be stabilized.

<Third Embodiment>
Next, an ignition timing control apparatus (hereinafter referred to as “third control apparatus”) according to a third embodiment of the present invention will be described. The third control device determines a case where it is not desirable to use the post-learning ignition timing as the final ignition timing based on the operating state of the engine 10, and an initial case where it is not desirable to use the post-learning ignition timing as the final ignition timing. It is different from the first control device only in that the ignition timing is set as the final ignition timing. Therefore, hereinafter, this difference will be mainly described.

  The CPU 71 of the third controller performs the ignition timing control routine shown in the flowchart of FIG. 7 instead of FIG. 4 every time the crank angle of a specific cylinder matches a predetermined crank angle (for example, BTDC 90 °) immediately before the start of the combustion stroke. Repeatedly executed. Also in FIG. 7, steps for performing the same processing as the steps shown in FIG. 4 are given the same reference numerals as those given for such steps in FIG. 4. A detailed description of these steps is omitted.

  When the crank angle of the specific cylinder coincides with the predetermined crank angle, the CPU 71 starts the processing of the routine of FIG. 7 from step 700 and performs the processing of step 405 to step 425, step 450 and step 455 described above. Thereby, the ignition timing control model is corrected by learning based on the previous combustion (step 425). Further, a post-learning ignition timing SAgk and an initial ignition timing SAin are calculated (step 450, step 455).

  Next, the CPU 71 proceeds to step 705, and determines whether or not the amount of change in the throttle valve opening TA per unit time (that is, the throttle valve opening changing speed) ΔTA is smaller than a predetermined throttle valve opening changing speed threshold ΔTAth. judge. The throttle valve opening TA is controlled by the CPU 71 and the throttle valve actuator 43a so as to increase as the operation amount Accp of the accelerator pedal 81 increases. In addition, the amount of change per unit time of the throttle valve opening TA is a routine executed every time a predetermined time Δt (not shown) is executed, and the throttle valve opening TA at the time of executing the routine and the previous execution of the routine are executed. It is obtained by obtaining the difference (ΔTA = TA−TAold) between the throttle valve opening TA (= TAold = throttle valve opening before Δt) at that time.

  Assume that the throttle valve opening change rate ΔTA is smaller than the throttle valve opening change rate threshold ΔTAth. In this case, since the operating state of the engine has not changed suddenly, a sufficient number of times of learning has been performed in the current operating region (learning region) and in the nearby operating region. Therefore, the post-learning ignition timing obtained by the post-learning ignition timing control model is a value close to an appropriate value (a highly reliable value).

  Accordingly, the CPU 71 determines “Yes” in step 705 and proceeds to step 465 to set the post-learning ignition timing SAgk obtained in step 450 as the final ignition timing SA (k). Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) set in step 465 (that is, the post-learning ignition timing SAgk). . Then, the CPU 71 proceeds to step 495 and once ends this routine.

  On the other hand, when the throttle valve opening change rate ΔTA is greater than the throttle valve opening change rate threshold ΔTAth (when the throttle valve opening change rate ΔTA is equal to or greater than the throttle valve opening change rate threshold ΔTAth), the operation region is relatively short. Since it changes in time, it is considered that learning of the ignition timing control model in the current operation region is not sufficient. In other words, since the learning does not sufficiently follow the change in the driving state, it is considered that the initial ignition timing SAin is a more reliable value than the post-learning ignition timing SAgk.

  Therefore, in this case, the CPU 71 determines “No” in step 705 and proceeds to step 475 to set the initial ignition timing SAin obtained in step 455 to the final ignition timing SA (k). Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) set in step 475 (that is, the initial ignition timing SAin). Then, the CPU 71 proceeds to step 495 and once ends this routine. The above is the operation of the third control device.

  FIG. 8 is a time chart for explaining the operation of the third control device during acceleration operation, which is transient operation. In FIG. 8, until time t1, the throttle valve opening change rate ΔTA indicated by the curve C1 is equal to or less than the throttle valve opening change rate threshold ΔTAth. Therefore, ignition is performed at the after-learning ignition timing SAgk indicated by the dashed curve C3. However, until time t1, learning is not significantly delayed with respect to changes in the operating state of the engine, so the after-learning ignition timing SAgk indicated by the dashed curve C3 is the initial ignition timing SAin indicated by the solid curve C2. And a value closer to the optimum value than the initial ignition timing SAin.

  After time t1, the throttle valve opening change speed ΔTA becomes larger than the throttle valve opening change speed threshold ΔTAth. In this case, the post-learning ignition timing SAgk gradually changes to an optimum value even after time t1. However, since the optimal ignition timing changes abruptly similar to the initial ignition timing SAin, if ignition is performed at the post-learning ignition timing SAgk, the ignition timing will advance too much, causing knocking and the like. Therefore, after the time t1, the third control device performs ignition at the initial ignition timing SAin instead of the after-learning ignition timing SAgk.

  After time t2, the throttle valve opening change rate ΔTA becomes equal to or less than the throttle valve opening change rate threshold ΔTAth. Since the learning is sufficiently advanced by this time, the post-learning ignition timing SAgk changes to a value close to the optimum value. Therefore, after the time t2, the third control device performs ignition at the after-learning ignition timing SAgk in which individual characteristics of the engine are reflected. As a result, before time t1 and after time t2, ignition is performed at a more optimal timing than when ignition is performed at the initial ignition timing SAin.

  Thus, according to the third control apparatus, when the load change rate (throttle valve opening change rate ΔTA) is larger than a predetermined load change rate threshold (throttle valve opening change rate threshold ΔTAth), the ignition timing by learning It is determined that the correction of the control model is not in time, and in this case, ignition timing control means for performing ignition at the initial ignition timing SAin instead of the after-learning ignition timing SAgk is provided. Thereby, it is possible to avoid the ignition timing from greatly deviating from an appropriate value during transient operation.

  In step 705, it is determined whether or not the change amount ΔTA of the throttle valve opening TA per unit time is smaller than a predetermined throttle valve opening change speed threshold ΔTAth. It may be configured to determine whether or not the absolute value of the change amount ΔTA per unit time of the opening degree TA is smaller than the throttle valve opening change rate threshold value ΔTAth. Further, in step 705, instead of the change amount per unit time of the throttle valve opening TA, the change amount per unit time of the operation amount Accp of the accelerator pedal 81 and the change amount per unit time of the load KL may be used. Good.

<Fourth embodiment>
Next, an ignition timing control apparatus (hereinafter referred to as “fourth control apparatus”) according to a fourth embodiment of the present invention will be described. The fourth control device is different from the first control device in that the post-learning ignition timing is limited based on the initial ignition timing, and the limited post-learning ignition timing is set as the final ignition timing. Therefore, hereinafter, this difference will be mainly described.

  The CPU 71 of the fourth controller performs the ignition timing control routine shown in the flowchart of FIG. 9 instead of FIG. 4 every time the crank angle of a specific cylinder matches a predetermined crank angle (for example, BTDC 90 °) immediately before the start of the combustion stroke. Repeatedly executed. Also in FIG. 9, steps for performing the same processing as the steps shown in FIG. 4 are denoted by the same reference numerals as those given for such steps in FIG. 4. A detailed description of these steps is omitted.

  When the crank angle of the specific cylinder coincides with the predetermined crank angle, the CPU 71 starts the processing of the routine of FIG. 9 from step 900 and performs the processing of step 405 to step 425, step 450, and step 455 described above. Thereby, the ignition timing control model is corrected by learning based on the previous combustion (step 425). Further, a post-learning ignition timing SAgk and an initial ignition timing SAin are calculated (step 450, step 455).

  Next, the CPU 71 proceeds to step 905 to determine whether or not the value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk is greater than a predetermined threshold (guard width) Dth (Dth> 0). In other words, in step 905, it is determined whether or not the post-learning ignition timing SAgk is a value further advanced than the ignition timing advanced by the threshold value Dth from the initial ignition timing SAin.

  Now, it is assumed that the value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk is larger than a predetermined threshold value Dth. As described above, the initial ignition timing control model is adapted in advance so that the 8 ° combustion ratio matches the target combustion ratio MFB8tgt. Therefore, the value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk is larger than the predetermined threshold value Dth. This is because learning is performed at a high frequency in an operation region different from the current operation region. This means that the post-learning ignition timing control model for the current operating region (and hence the post-learning ignition timing for the current operating region) is inappropriate. Based on this viewpoint, the fourth control device limits the post-learning ignition timing based on the initial ignition timing.

  More specifically, the CPU 71 makes a “Yes” determination at step 905 to proceed to step 910 where a value obtained by adding a threshold Dth to the initial ignition timing SAin (an ignition timing obtained by advancing the initial ignition timing SAin by the threshold Dth). ) As the final ignition timing SA (k). The ignition timing obtained by adding the threshold value Dth to the initial ignition timing SAin is also referred to as a limited ignition timing. Next, the CPU 71 proceeds to step 915 to set the ignition timing deviation amount ΔSA (k) to the threshold value Dth, and proceeds to step 920 to execute learning of the ignition timing control model again. That is, in step 920, the CPU 71 calculates (determines) the coefficients θ0, θ1, θ2, and θ3 that minimize the ignition timing deviation amount ΔSA (k) based on the sequential least square method. At this time, the CPU 71 uses the engine speed NE (k) and the load KL (k) for calculation.

  Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) determined in step 910. Then, the CPU 71 proceeds to step 995 to end the present routine tentatively. As a result, actual ignition is executed at the limited ignition timing obtained by advancing the initial ignition timing SAin by the threshold value Dth.

  On the other hand, when the processing of step 455 is completed, the value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk is smaller than the predetermined threshold Dth, but the value obtained by subtracting the post-learning ignition timing SAgk from the initial ignition timing SAin Assume that it is greater than a predetermined threshold Dth. Also in this case, as a result of learning being performed frequently in an operation region different from the current operation region, a post-learning ignition timing control model for the current operation region (and hence a post-learning ignition timing for the current operation region) is obtained. It means that it is inappropriate. Based on this viewpoint, the fourth control device limits the post-learning ignition timing based on the initial ignition timing.

  That is, the CPU 71 makes a “No” determination at step 905 and proceeds to step 925 to determine whether or not the value obtained by subtracting the post-learning ignition timing SAgk from the initial ignition timing SAin is greater than a predetermined threshold Dth. In other words, it is determined whether or not the post-learning ignition timing SAgk is a value further retarded than the ignition timing obtained by retarding the initial ignition timing SAin by the threshold value Dth. According to the above assumption, this condition is satisfied. Accordingly, the CPU 71 makes a “Yes” determination at step 925 and proceeds to step 930 to finalize the value obtained by subtracting the threshold Dth from the initial ignition timing SAin (ignition timing obtained by retarding the initial ignition timing SAin by the threshold Dth). Determined as time SA (k). The ignition timing obtained by subtracting the threshold value Dth from the initial ignition timing SAin is also referred to as a limited ignition timing. Then, the CPU 71 proceeds to step 935, sets the ignition timing deviation amount ΔSA (k) to a value (−Dth) obtained by inverting the sign of the threshold value Dth, and proceeds to step 920 to execute learning of the ignition timing control model.

  Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) determined in step 930. Then, the CPU 71 proceeds to step 995 to end the present routine tentatively. As a result, actual ignition is executed at the limited ignition timing obtained by retarding the initial ignition timing SAin by the threshold value Dth.

  On the other hand, when the absolute value of the difference between the post-learning ignition timing SAgk and the initial ignition timing SAin is smaller than Dth (below Dth) at the time when the processing of step 455 is finished, post-learning ignition timing control for the current operating region. The model (and hence the post-learning ignition timing for the current operating region) is not considered inappropriate.

  Therefore, in this case, the CPU 71 determines “No” in both step 905 and step 925 and proceeds to step 940 to set the after-learning ignition timing SAgk as the final ignition timing SA (k). Next, the CPU 71 proceeds to step 470 and sends an instruction signal to the igniter 38 so that ignition is performed at the final ignition timing SA (k) determined in step 940. Then, the CPU 71 proceeds to step 995 to end this routine once. As a result, actual ignition is executed at the after-learning ignition timing SAgk.

  As described above, the fourth control device, when the difference between the after-learning ignition timing SAgk and the initial ignition timing SAin (the magnitude of the difference, that is, the meaning of the absolute value of the difference) is smaller than the predetermined threshold Dth. The post-learning ignition timing SAgk is determined as the final ignition timing SA (k), and when the difference is greater than the predetermined threshold SAgk, the limited ignition timing (initial ignition timing SAin is determined by the threshold Dth) determined based on the initial ignition timing SAin. The ignition timing advanced or the initial ignition timing SAin retarded by the threshold value Dth) is determined as the final ignition timing SA (k), and ignition is performed at the determined final ignition timing SA (k). Ignition timing control means is provided. As a result, since the ignition timing is not set to the inappropriate after-learning ignition timing SAgk, it is possible to avoid the ignition timing from greatly deviating from the appropriate value.

  The limited ignition timing is set to an ignition timing obtained by advancing the initial ignition timing SAin by a first predetermined amount (Dth1> 0) when the after-learning ignition timing SAgk is on the advance side of the initial ignition timing SAin. When the post-learning ignition timing SAgk is retarded from the initial ignition timing SAin, the initial ignition timing SAin is retarded by a second predetermined amount (Dth2> 0) different from the first predetermined amount (Dth1). It may be set. In order to suppress knocking due to excessive advance of the ignition timing, it is desirable that the first predetermined amount Dth1 is smaller than the second predetermined amount Dth2. Further, the predetermined threshold value Dth may be constant, or may be a value that changes in accordance with the engine operating state quantity. In particular, in order to avoid occurrence of knocking on the high load side, the predetermined threshold value Dth may be set so as to decrease as the engine load (for example, load KL) increases.

  Further, when the ignition is executed at the limited ignition timing, the fourth control device applies the engine rotational speed NE (k) and the engine load KL (k) at that time to the post-learning ignition timing control model at that time. Then, the coefficients θ0 to θ3 are corrected so that the post-learning ignition timing control model calculates the ignition timing of the same limited ignition timing (relearning is performed). Therefore, the fact that the engine is limited (and the above-mentioned limited ignition timing) can be reflected in the post-learning ignition timing control model by learning. As a result, the post-learning ignition timing control model can be brought closer to a more appropriate model.

  Next, modified examples of the first to third control devices will be described. In these modified examples, re-learning similar to Step 915, Step 935, and Step 920 of FIG. 9 executed by the fourth control device is performed.

(Modification of the first control device)
In the modified example of the first control device, the processing of Step 1005 and Step 1010 shown in FIG. 10 is performed between Step 475 and Step 470 of the routine shown in FIG. More specifically, the CPU 71 determines “No” in step 460 when the learning ratio (= Cx, y / CN) obtained by dividing the learning counter for each driving region Cx, y by the total learning number counter CN is smaller than the threshold value Th. In step 475, the initial ignition timing SAin obtained in step 455 is set as the final ignition timing SA (k). Thereby, in the current operating region to which the current operating state represented by the engine speed NE (k) and the load KL (k) belongs, a more reliable initial ignition timing SAin is adopted instead of the post-learning ignition timing SAgk. Is done.

  Next, the CPU 71 proceeds to step 1005 to set “a value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk (SAgk−SAin)” as the ignition timing deviation amount ΔSA (k). Then, the CPU 71 proceeds to step 1010, and again executes learning of the ignition timing control model as in step 920 of FIG. Specifically, the CPU 71 calculates (determines) the coefficients θ0, θ1, θ2, and θ3 that minimize the ignition timing deviation amount ΔSA (k) (= SAgk−SAin) based on the successive least square method. At this time, the CPU 71 uses the engine speed NE (k) and the load KL (k) for calculation. Thereafter, the CPU 71 proceeds to step 470 in FIG. 4 to perform an ignition execution process.

  As a result, the post-learning ignition timing control model in the current operation region can be brought closer to a more appropriate model.

(Modification of the second control device)
In the modified example of the second control device, the processing of Step 1105 and Step 1110 shown in FIG. 11 is performed between Step 605 and Step 470 of the routine shown in FIG. More specifically, the CPU 71 proceeds to step 1105 in FIG. 11 following step 605 in FIG. 6, and determines the ignition timing deviation amount ΔSA (k) from the “learned ignition timing SAgk in step 605 in FIG. 6. A value obtained by subtracting the final ignition timing SA (k) is set. As described above, the final ignition timing SA (k) is an ignition timing obtained by weighting and averaging the after-learning ignition timing SAgk and the initial ignition timing SAin using the learning ratio (= Cx, y / CN).

  Then, the CPU 71 proceeds to step 1110 and executes learning of the ignition timing control model again as in step 920 of FIG. Specifically, the CPU 71 calculates (determines) the coefficients θ0, θ1, θ2, and θ3 that minimize the ignition timing deviation amount ΔSA (k) (= SAgk−SA (k)) based on the sequential least square method. To do. At this time, the CPU 71 uses the engine speed NE (k) and the load KL (k) for calculation. Thereafter, the CPU 71 proceeds to step 470 in FIG. 6 to perform an ignition execution process.

  As a result, the post-learning ignition timing control model in the current operation region can be brought closer to a more appropriate model.

(Modification of the third control device)
In the modified example of the third control device, the processing of Step 1205 and Step 1210 shown in FIG. 12 is performed between Step 475 and Step 470 of the routine shown in FIG. More specifically, if the throttle valve opening change rate ΔTA is greater than the throttle valve opening change rate threshold ΔTAth, the CPU 71 makes a “No” determination at step 705 to proceed to step 475, where the final ignition timing SA ( k) is set to the initial ignition timing SAin obtained in step 455 of FIG. Thereby, in the current operating region to which the current operating state represented by the engine speed NE (k) and the load KL (k) belongs, a more reliable initial ignition timing SAin is adopted instead of the post-learning ignition timing SAgk. Is done.

  Next, the CPU 71 proceeds to step 1205 to set “a value obtained by subtracting the initial ignition timing SAin from the post-learning ignition timing SAgk (SAgk−SAin)” as the ignition timing deviation amount ΔSA (k). Then, the CPU 71 proceeds to step 1210 and performs learning of the ignition timing control model again as in step 920 of FIG. Specifically, the CPU 71 calculates (determines) the coefficients θ0, θ1, θ2, and θ3 that minimize the ignition timing deviation amount ΔSA (k) (= SAgk−SAin) based on the successive least square method. At this time, the CPU 71 uses the engine speed NE (k) and the load KL (k) for calculation. Thereafter, the CPU 71 proceeds to step 470 in FIG. 4 to perform an ignition execution process.

  As a result, the post-learning ignition timing control model in the current operation region can be brought closer to a more appropriate model.

  As is apparent from the above, each of the modified examples of the first to third control devices and the fourth control device are configured so that the final ignition timing SA (k) and the post-learning ignition timing SAgk are different from each other. Model learning means (relearning means) for performing learning (relearning) for correcting the function representing the ignition timing control model so that the difference between the final ignition timing SA (k) and the post-learning ignition timing SAgk is reduced. You can also say.

  As described above, according to each embodiment of the ignition timing control according to the present invention, since the inappropriate post-learning ignition timing SAgk is not used as the ignition timing, knocking, occurrence of unstable combustion state, and the like are avoided. be able to.

  The present invention is not limited to the above embodiment, and various modifications can be employed within the scope of the present invention. For example, in each of the above embodiments, learning of the ignition timing control model is performed by the sequential least square method, but learning may be performed using other methods. The learning of the ignition timing control model has been performed by correcting all the coefficients θ0 to θ3. However, even if the learning is executed so as to correct only some of the coefficients θ0 to θ3. Good. In addition, when the ignition timing control model has adaptation parameters other than coefficients θ0 to θ3, the initial ignition timing control model is a model in which such adaptation parameters are adapted, and the post-learning ignition timing control model is such a model. The adaptation parameter may be corrected by learning.

  In each of the above embodiments, the combustion rate MFB (accordingly, the indicated rate of heat quantity Qsum / Qtotal) is obtained based on the in-cylinder pressure. (See Japanese Patent No. 9720). Further, the combustion state index value is not limited to the combustion ratio MFB at a predetermined crank angle, and may be, for example, the in-cylinder pressure and / or the change rate of the in-cylinder pressure at a predetermined crank angle.

1 is a schematic view of an internal combustion engine to which an ignition timing control device according to a first embodiment of the present invention is applied. It is the graph which showed the relationship between ignition timing, an 8 degree combustion ratio, and the generated torque of an engine. FIG. 2 is a functional block diagram of the electric control device (ignition timing control device) shown in FIG. 1. It is the flowchart which showed the routine which CPU shown in FIG. 1 performs. It is the figure which showed an example of the learning area | region of an ignition timing control model. It is the flowchart which showed the routine which CPU of the ignition timing control apparatus which concerns on 2nd Embodiment of this invention performs. It is the flowchart which showed the routine which CPU of the ignition timing control apparatus which concerns on 3rd Embodiment of this invention performs. It is a time chart for demonstrating the action | operation of the ignition timing control apparatus which concerns on 3rd Embodiment at the time of acceleration operation. It is the flowchart which showed the routine which CPU of the ignition timing control apparatus which concerns on 4th Embodiment of this invention performs. It is the flowchart which showed a part of routine which CPU of the ignition timing control apparatus which concerns on the modification of 1st Embodiment of this invention performs. It is the flowchart which showed a part of routine which CPU of the ignition timing control apparatus which concerns on the modification of 2nd Embodiment of this invention performs. It is the flowchart which showed a part of routine which CPU of the ignition timing control apparatus which concerns on the modification of 3rd Embodiment of this invention performs. It is the figure which showed the ignition timing calculated by the ignition timing control model when learning is performed without a bias with respect to all the operation areas. It is the figure which showed the ignition timing calculated by the ignition timing control model after learning was performed only in a certain specific operation area | region. It is a graph showing the frequency according to the driving | operation area | region of learning of the ignition timing control model after performing a predetermined driving | operation.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Internal combustion engine, 25 ... Combustion chamber, 37 ... Spark plug, 38 ... Igniter, 43 ... Throttle valve, 65 ... In-cylinder pressure sensor, 69 ... Accelerator opening sensor, 70 ... Electric control device, 71 ... CPU, A1 ... Operation State quantity acquisition unit, A2 ... target value setting unit, A3 ... actual combustion ratio MFB8 calculation unit, A4 ... ignition timing deviation amount calculation unit, A5 ... delay unit, A6 ... model learning unit, A7 ... ignition timing control unit.

Claims (6)

An ignition timing control model expressed by a function for calculating the ignition timing of the engine with the operating state quantity representing the operating state of the internal combustion engine as a variable, and a combustion state index value indicating the combustion state of the engine is a predetermined target combustion state An ignition timing control device for an internal combustion engine comprising an initial ignition timing control model adapted in advance to match a target combustion state index value indicating
Driving state quantity acquisition means for acquiring the actual driving state quantity as an actual driving state quantity;
Initial ignition timing calculation means for calculating an initial ignition timing by applying the acquired actual operating state quantity to the initial ignition timing control model;
Combustion state index value acquisition means for acquiring the actual combustion state index value as an actual combustion state index value;
Learning to sequentially correct the function representing the initial ignition timing control model so that the difference between the acquired actual combustion state index value and the target combustion state index value becomes smaller when the actual combustion state index value is acquired. Model learning means to perform,
A post-learning ignition timing calculating means for calculating a post-learning ignition timing by applying the acquired actual operating state quantity to the ignition timing control model learned by the model learning means;
Total learning number acquisition means for acquiring a total learning number by accumulating the number of times of learning by the model learning means;
The engine operating region is divided into a plurality of operating regions in advance, and the number of times of learning by the model learning means is the same as the plurality of operations to which the operating state indicated by the actual operating state quantity when the learning is performed belongs. A learning area-by-driving area acquisition means for acquiring the learning area-by-driving area count by accumulating corresponding to one driving area in the area;
The number of learnings for each driving region acquired corresponding to the current driving region that is one of the plurality of driving regions to which the driving state indicated by the actual driving state amount at the present time belongs, and the total learning And a learning ratio that is a value corresponding to the ratio of the number of times and a final ignition timing that reflects the initial ignition timing and the post-learning ignition timing according to the learning ratio. Ignition timing control means for performing ignition at
Ignition timing control device. The ignition timing control device for an internal combustion engine according to claim 1,
The combustion state index value is an actual combustion ratio when the crank angle of the engine is a predetermined crank angle, and the target combustion state index value is a combustion when the crank angle of the engine is the predetermined crank angle. Ignition timing control device that is the target value of the ratio. An ignition timing control device for an internal combustion engine according to claim 1 or 2,
The ignition timing control means includes
When the learning ratio for the current operating region is larger than a predetermined ratio, the post-learning ignition timing is determined as the final ignition timing, and when the learning ratio is smaller than the predetermined ratio, the initial ignition timing is set as the final ignition timing. An ignition timing control device configured to determine. An ignition timing control device for an internal combustion engine according to claim 1 or 2,
The ignition timing control means includes
The greater the learning ratio for the current operating region, the greater the weight of the post-learning ignition timing and the smaller the weight of the initial ignition timing, so that the same learning ratio is set for the post-learning ignition timing and the initial ignition timing. An ignition timing control device configured to obtain the final ignition timing by performing a corresponding weighting. An ignition timing control model expressed by a function for calculating the ignition timing of the engine with the operating state quantity representing the operating state of the internal combustion engine as a variable, and a combustion state index value indicating the combustion state of the engine is a predetermined target combustion state An ignition timing control device for an internal combustion engine comprising an initial ignition timing control model adapted in advance to match a target combustion state index value indicating
Driving state quantity acquisition means for acquiring the actual driving state quantity as an actual driving state quantity;
Initial ignition timing calculation means for calculating an initial ignition timing by applying the acquired actual operating state quantity to the initial ignition timing control model;
Combustion state index value acquisition means for acquiring the actual combustion state index value as an actual combustion state index value;
Learning to sequentially correct the function representing the initial ignition timing control model so that the difference between the acquired actual combustion state index value and the target combustion state index value becomes smaller when the actual combustion state index value is acquired. Model learning means to perform,
A post-learning ignition timing calculating means for calculating a post-learning ignition timing by applying the acquired actual operating state quantity to the ignition timing control model learned by the model learning means;
Load change speed acquisition means for acquiring the load change speed of the engine;
When the acquired load change speed is smaller than a predetermined load change speed threshold, the post-learning ignition timing is determined as the final ignition timing, and when the acquired load change speed is larger than the load change speed threshold. An ignition timing control means for determining the initial ignition timing as a final ignition timing and executing ignition at the determined final ignition timing;
Ignition timing control device. An ignition timing control model expressed by a function for calculating the ignition timing of the engine with the operating state quantity representing the operating state of the internal combustion engine as a variable, and a combustion state index value indicating the combustion state of the engine is a predetermined target combustion state An ignition timing control device for an internal combustion engine comprising an initial ignition timing control model adapted in advance to match a target combustion state index value indicating
Driving state quantity acquisition means for acquiring the actual driving state quantity as an actual driving state quantity;
Initial ignition timing calculation means for calculating an initial ignition timing by applying the acquired actual operating state quantity to the initial ignition timing control model;
Combustion state index value acquisition means for acquiring the actual combustion state index value as an actual combustion state index value;
Learning to sequentially correct the function representing the initial ignition timing control model so that the difference between the acquired actual combustion state index value and the target combustion state index value becomes smaller when the actual combustion state index value is acquired. Model learning means to perform,
A post-learning ignition timing calculating means for calculating a post-learning ignition timing by applying the acquired actual operating state quantity to the ignition timing control model learned by the model learning means;
When the difference between the post-learning ignition timing and the initial ignition timing calculation is smaller than a predetermined threshold, the post-learning ignition timing is determined as a final ignition timing, and when the difference is larger than the predetermined threshold, it is determined based on the initial ignition timing. Ignition timing control means for determining the limited ignition timing to be determined as the final ignition timing, and executing ignition at the determined final ignition timing;
Ignition timing control device.

Images