JPH10184438A - Air amount detecting device for engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To further improve calculation precision of a equivalent signal to a cylinder air amount by solving the physical model of an intake system considering the time change of a cylinder intake volume. SOLUTION: A calculating means 3 calculates a equivalent signal to an intake air amount per unit rotation from the output of an air flow meter 2 and a calculating means 4 calculates first and second weighted mean factors Fload 1 and Fload 2 by considering the time change of a cylinder intake volume (intending for example, a sum of the Fload 1 and the Fload 2 is 1 during steady operation and not 1 during transient operation). A sum of a value obtained by multiplying the equivalent signal to an intake air amount per unit rotation by the fist weighted means factor Fload 1 and a value obtained by multiplying a equivalent signal to a preceding cylinder air amount by the second weighted means factor Fload 2 is calculated as a equivalent signal to a present cylinder air amount by a calculating means 5.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for detecting an air amount of an engine, and more particularly to a device for correcting a response delay in a transient state.

[0002]

2. Description of the Related Art In an L-jetronic type fuel injection device provided with an injector near the cylinder of each cylinder, an intake air amount Qa detected by an air flow meter and an engine speed detected by a crank angle sensor and the like. The basic injection pulse width Tp is calculated using Ne and Tp = K · Q / Ne (K is a constant). By performing a weighted average on this Tp, the pulse width (cylinder equivalent to the cylinder air amount) is calculated. Injection pulse width proportional to the amount of air in the vicinity) Avt
There is a method for obtaining p (see Japanese Patent Application No. 1-248677).

[0003] The weighted average is required because the amount of air at the position of the air flow meter includes an overshoot due to the pressure change of the internal volume of the manifold during transition. This is because it is necessary to remove this overshoot. In other words, by repeatedly executing formula (1) described later at a constant calculation cycle, the cylinder air amount which changes transiently is predicted, and the fuel amount is to be given in proportion to the predicted air amount. It is.

[0004]

In this case, even if the operating conditions such as the load and the number of revolutions of the engine are different, the engine speed Ne and the engine load correspond so that the cylinder air amount can be accurately predicted. If a map of the weighted average coefficient is created by using the amount (for example, the throttle valve opening TVO) as a parameter and the map value is adapted on an actual machine, the time required for the map adaptation is prolonged, and the number of development man-hours increases. It is necessary to adapt to each.

When a weighted average is applied to Tp, which causes a large response delay, a pulse width A corresponding to the cylinder air amount is obtained.
The phase of vtp does not match the phase of the actual cylinder air amount, and during acceleration, Avtp is accompanied by a delay in rising and a subsequent overresponse (overshoot). In the section with such a rise delay and overshoot, the air-fuel ratio deviates from the target value, and the drivability and emission deteriorate.

In a specific delay system for calculating the air amount,
Low-pass filter for removing ignition noise (time constant is 20m
Due to the delay caused by s) and the like, the delay system is a multiplexed series connection such as a first-order delay, so that the rising delay and overshoot occurring in Avtp are further increased.

Therefore, a calculation formula Fload1 = 1 / ｛(120 · Vm) / (Ve · η · Ne determined from the physical model of the intake system.
.DELTA.t) + 1 @ Fload2 = 1-Fload1 where Fload1: first weighted average coefficient Fload2: second weighted average coefficient Vm: manifold volume Ve: displacement η: fresh air ratio in cylinder Ne: engine speed Δt: The two weighted average coefficients Fload1 and Fload2 are obtained from the calculation time interval
By using these weighted average coefficients Fload1 and Fload2 to calculate the pulse width Avtp equivalent to the cylinder air amount using the following equation, Avtp = Tpflat · Fload1 + old Avtp · Fload2, the detection man-hours during the transition are not reduced and the number of development steps is reduced. (Japanese Patent Laid-Open No. 4-237)
No. 856).

However, although the cylinder intake volume Vc actually changes temporally with respect to the transient change, the above-described conventional apparatus solves the temporal change of the cylinder intake volume Vc when solving the physical model of the intake system. Because it does n’t take change into account,
The temporal change of Vc has occurred as a calculation error of the signal corresponding to the cylinder air amount.

Accordingly, an object of the present invention is to solve the physical model of the intake system in consideration of the temporal change of the intake volume in the cylinder, thereby further improving the calculation accuracy of the signal corresponding to the cylinder air amount.

[0010]

According to a first aspect of the present invention, as shown in FIG. 14, there is provided an air flow meter 2 for producing an output corresponding to the amount of air upstream of an intake throttle valve 1, and an output per unit rotation based on the output of the air flow meter. Means 3 for calculating a signal corresponding to the intake air amount (for example, basic injection pulse width Tp), and calculating a first weighted average coefficient Fload1 and a second weighted average coefficient Fload2 in consideration of a temporal change of the intake volume in the cylinder. (For example, means 4 for calculating so that the sum of Fload1 and Fload2 is 1 in a steady state and not 1 in a transient state (during acceleration and deceleration)) and a signal corresponding to the amount of intake air per unit rotation. A value obtained by multiplying the weighted average coefficient Fload1 by 1 and a signal corresponding to the previous cylinder air amount (for example, the previous Avtp which is the previous value of the pulse width corresponding to the cylinder air amount) is multiplied by the second weighted average coefficient Fload2. And means 5 for calculating the sum of the current cylinder air amount equivalent signal (e.g. the cylinder air amount equivalent pulse width AvTp) with.

According to a second aspect, in the first aspect, the two weighted average coefficients Fload1 and Fload2 are calculated based on a change rate β of the in-cylinder fresh air ratio η (n).

In a third aspect, in the second aspect, the rate of change of the fresh air ratio in the cylinder is β = η (n-1) / η (n), where β is the rate of change of the fresh air rate in the cylinder η (N): In-cylinder fresh air ratio η (n-1): Calculated by the formula of the previous value of in-cylinder fresh air ratio.

According to a fourth aspect of the present invention, in the first aspect, the two weighted average coefficients are expressed by: Fload1 = 1 / (1 + Kfl · β) Fload2 = Kfl / (1 + Kfl · β) Kfl = (120 · Vm) / (Ve · η (n) · Ne · Δ
t) β = η (n−1) / η (n) where Fload1: first weighted average coefficient Fload2: second weighted average coefficient Kfl: coefficient β: rate of change of fresh air ratio in cylinder Vm: manifold volume Ve: Displacement amount η (n): In-cylinder fresh air ratio η (n-1): Previous value of the in-cylinder fresh air ratio Ne: Engine speed Δt: Calculation time interval

According to a fifth aspect of the present invention, in the first aspect, the two weighted average coefficients are represented by Fload1 = 1 / Kfl Fload2 = (1-Fload1) / β Kfl = (120 · Vm) / (Ve · η (n)・ Ne ・ Δ
t) β = η (n−1) / η (n) where Fload1: first weighted average coefficient Fload2: second weighted average coefficient Kfl: coefficient β: rate of change of fresh air ratio in cylinder Vm: manifold volume Ve: Displacement amount η (n): In-cylinder fresh air ratio η (n-1): Previous value of the in-cylinder fresh air ratio Ne: Engine speed Δt: Calculation time interval

According to a sixth aspect of the present invention, as shown in FIG. 15, the time constant of the low-pass filter 6 connected between the air flow meter 2 and the intake air amount signal calculating means 3 is at least 10 ms.
Within.

[0016]

According to the first aspect of the invention, in consideration of a temporal change in the intake air volume in the cylinder, for example, the sum of the first weighted average coefficient and the second weighted average coefficient is 1 in a steady state, and is 1 in a transient state. Since each weighted average coefficient is calculated so as not to be 1, the rising delay immediately after acceleration (falling delay in the initial stage of deceleration during deceleration) and the overshoot in the latter period of acceleration are significantly improved as in the conventional device. In the case of the present invention, the change in the actual air amount in the cylinder at the time of transition can be made closer to that of the conventional device.

In the first invention, since the rate of change of the fresh air ratio in the cylinder is constant, if the rate of change differs from the rate of change at this time, the accuracy of calculating the cylinder air amount deteriorates. On the other hand, in the second invention, two weighted average coefficients are calculated based on the rate of change of the in-cylinder fresh air rate in response to the change rate of the in-cylinder fresh air rate changing when the degree of the transient is different. Since the calculation is performed, it is possible to approach the actual change in the air amount in the cylinder during the transition even if the degree of the transition is different.

According to the respective formulas for calculating the coefficient Kfl and the change rate of the cylinder fresh air ratio as in the fourth invention, the cylinder fresh air ratio at that time, the previous value of the cylinder fresh air ratio and the engine speed By simply substituting numbers, two weighted average coefficients are calculated. In other words, when preparing two weighted average coefficients, it is only necessary to store the engine specifications (Vm, Ve) of the applicable model in a memory in advance, so that a map of the two weighted average coefficients is created. This is unnecessary, and the number of steps for matching can be reduced. Even if the applicable model is different, it is only necessary to change the engine specifications accordingly.

Further, since the equation is a theoretical equation, the actual cylinder air amount can be accurately predicted in calculating the equation, as long as the delay generated in the delay system assumed as a premise is small. .

In the fifth aspect, the calculation of the two weighted average coefficients is easier than in the fourth aspect.

On the other hand, when the air-fuel ratio control is performed using the signal corresponding to the cylinder air amount, as the time constant of the low-pass filter increases, the amount of deviation in the lean direction increases. When the time constant is set to at least 10 ms in the present invention, the amount of deviation in the lean direction is reduced, and the lean spike during acceleration is further improved.

[0022]

FIG. 1 is a system diagram of an embodiment.
The air amount detecting device according to the present invention is applied to an L-jetronic fuel injection device. In the figure, the air taken into the intake passage 11 from the air cleaner 10 is:
The fuel is mixed with fuel injected from an injector 13 provided in the intake port 12 and guided into a cylinder 14 of the engine.
The gas that burns in the cylinder with the help of ignition sparks,
The work of pushing down the piston is performed, and after the work, the combustion gas passes through the exhaust passage 15 and emits harmful three components (CO, HC, N).
Ox) is led to the catalytic converter 16 for purifying Ox).

When the accelerator pedal is depressed, the intake throttle valve 17 associated therewith opens, and the amount of air flowing into the cylinder 14 increases. This change in the amount of air changes the signal of the hot-wire air flow meter 18 interposed upstream of the throttle valve 17.

The air flow meter 18 comprises a sensor section and a control circuit (not shown). In the sensor section, a heating wire is wound in a coil shape around a bobbin-shaped ceramic, and the entire sensor section is coated with glass. It is covered with. In the control circuit, a bridge circuit is formed between the heating wire. In this case, the hot wire is cooled as the air amount increases. Therefore, if the supply current to the hot wire is increased so that the temperature of the hot wire becomes constant, the supply current value corresponds to the air amount. This current value is converted to a voltage value and output.

The control unit 25 mainly composed of a microcomputer has a sensor 19 for detecting the throttle valve opening TVO and a crank angle sensor for generating a predetermined pulse signal in synchronization with the crankshaft (built into the distributor 20). ) 21, water temperature sensor 22, oxygen sensor 2
3, the control unit 25 determines the fuel injection amount from the injector 13 in accordance with the cylinder air amount obtained based on the air flow meter output.

In this case, the amount of air flowing near the cylinder and the amount of air flowing through the position of the air flow meter are different during a transition due to the manifold volume, so that the fuel injection pulse width given to the injector at the injector position, that is, The basic injection pulse width Tp [ms], which must be given in proportion to the amount of air flowing near the cylinder and is calculated from the air flow meter output and the engine speed Ne,
The first weighted average coefficient Fload1, the second weighted average coefficient Flo
Using ad2 (= 1−Fload1), the cylinder air amount-equivalent pulse width Avtp [ms] is calculated by the following equation (details will be described later): Avtp = Tpflat · Fload1 + old Avtp · Fload2 (1) where oldtp: Avtp: Avtp Calculated using the previous value.

Thus, the pulse width A corresponding to the cylinder air amount
After obtaining vtp, the final fuel injection pulse width Ti [ms] is determined by the following equation, and fuel injection is performed.

Ti = Avtp × Co × α + Ts (2) Expression (2) is an expression for simultaneous injection of all cylinders (once for each revolution of the engine, simultaneous injection for each cylinder). Ti = Avtp × Co × α × 2 + Ts (injection once for each rotation according to the ignition order of each cylinder) (3)

However, in equations (2) and (3), Co
Is the sum of 1 and various correction coefficients, α is the air-fuel ratio feedback correction coefficient calculated based on the oxygen sensor output, and Ts is the invalid pulse width of the injector 13.

Now, when calculating the two weighted average coefficients Fload1 and Fload2 of the above equation (1) using an equation determined by a physical model of the intake system, if the temporal change of the intake volume in the cylinder is not taken into consideration, the cylinder internal The temporal change of the intake volume occurs as an error in the calculation of the cylinder air amount equivalent signal.

In order to cope with this, in the present invention, two weighted average coefficients Fload1 and Fload2 are obtained by solving a physical model of the intake system in consideration of the temporal change of the intake volume in the cylinder. In the calculation, it is assumed that the delay in the delay system for calculating the air amount is very small.

In the physical model of the intake system shown in FIG. 3, the amount of air charged to the manifold is represented by Gam (t) [g /
S] and the cylinder air amount is Gacy (t) [g / s], Gam (t) = Pm · Vm / (R · Tm) (11) Gacy (t) = Pc · Vc / (R · Tc) (12) Pc / Tc ≒ Pm / Tm (13) However, (t) added to Gam and Gacy indicates that it is a function of time t. (12)
The formula is a formula introduced as a cylinder suction model.

Here, the intake air volume in the cylinder of the equation (12) is a function of time (therefore, the fresh air ratio η [%] in the cylinder)
Vc (t) = Ve · η (t) · Ne / 120 (14) When the cylinder intake volume Vc (t) [cc] is given by the following equation, equation (12) is obtained. Gacy (t) = Pc.Vc (t) / (R.Tc) (12a).

Substituting equation (13) into equation (11), (12)
a) Using the equation, Gam (t) = (Vm / R) ・ (Pm / Tm) ≒ (Vm / R) ・ (Pc / Tc) (∵from equation (13)) = (Vm / R ) · (1 / Tc) · (R · Tc / Vc (t)) · Gacy (t) = (Vm / Vc (t)) · Gacy (t) (15)

The air amount of the throttle valve portion is represented by Gath (t) [g / s].
If (this is equal to the flow rate through the air flow meter), the difference between Gath (t) and Gacy (t) is the amount of air charged to the manifold. Therefore, the change at the transient time is Gath (t) −Gacy (t) = dGam / dt (16) Equation (16) is a manifold filling model.

The expression (15) is substituted into the right side of the expression (16). Gath (t) −Gacy (t) = Vm · d [Gacy (t) / Vc (t)] / dt = (Vm / Vc (t)) · dGacy (t) / dt + Vm · Gacy (t) · d [1 / Vc (t)] / dt = [120 · Vm / (Ve · η (t) · Ne)] · dGacy (t) / dt−Gacy (t) · [120 · Vm / (Ve · η ( t) · Ne)] · (1 / η (t)) · dη (t) / dt (17)

In this way, a continuous value equation is obtained.

The equation different from that of the prior application apparatus is the above (14),
Expressions (15) and (16). Vc (t) in equation (15)
Is treated not as a function of time but as a constant, and the differentiation of equation (16) is performed, the equation of the prior application is obtained.

The meanings of the symbols not explained in the above equations (11) to (17) are as follows. Pm: Manifold pressure Vm: Manifold volume [cc] Tm: Manifold air temperature [K] Pc: Cylinder pressure Tc: Cylinder air temperature [K] Ve: Displacement [cc] Ne: Engine speed [rpm] R: gas constant

Here, in order to convert the continuous value equation (17) into a discrete value equation, Gacy (t) → Gacy (n) η (t) → η (n) dGacy ( t) / dt → ΔGacy (n) / Δt dη (t) / dt → Δη (n) / Δt where Δt: sample interval (calculation time interval) n: number of samples Gath (n) −Gacy (n) = ΔGacy (n) · [120 · Vm / (Ve · η (n) · Ne · Δt)] − Gacy (n) · [120 · Vm / (Ve · η (n ) · Ne · Δt)] · (Δη (n) / η (n)) (18)

Here, ΔGacy (n) = Gacy (n) −Gacy (n−1) (19) Δη (n) = η (n) −η (n−1) (20) Is substituted into equation (18), then (18)
The formula is as follows: Gath (n) −Gacy (n) = Kfl · (Gacy (n) −Gacy (n−1)) − Kfl · Gacy (n) (1-β) = Kfl · β · Gacy (n) −Kfl · Gacy (N-1) ... (21)

However, in the equation (21), the coefficient Kfl and the change rate β of the fresh air ratio in the cylinder are Kfl [absolute number] = 120 · Vm / (Ve · η (n) · Ne · Δt) (22) = VOLR # / (η (n) · Ne) (23) where VOLR #: coefficient according to (manifold volume / displacement amount) β [absolute number] = η (n−1) / η (n) ( 24) is the value defined by

Therefore, the cylinder air is calculated by the following equation (25): Gacy (n) = Gath (n) / (1 + Kfl · β) + Gacy (n−1) · Kfl / (1 + Kfl · β) The quantity Gacy (n) could be expressed.

When this equation (25) is compared with the equation Avtp = Tpflat · Fload1 + old Avtp · Fload2 for calculating the pulse width equivalent to the cylinder air amount, the pulse width Avtp equivalent to the cylinder air amount becomes the cylinder air amount Gacy (n). Tpflat is the throttle valve air amount Gath (n), and old Avtp, the previous value of the pulse width corresponding to the cylinder air amount, is Gacy, the previous value of the cylinder air amount.
(N-1).

Here, since the cylinder air amount Gacy (n) and the pulse width Avtp corresponding to the cylinder air amount must have the same response, the two weighted average coefficients Fload1, Fload
2 is represented by Fload1 = 1 / (1 + Kfl · β) (26) Fload2 = Kfl / (1 + Kfl · β) (27)

Equations (26) and (27) show that the two weighted average coefficients Fload1 and Fload2 can be obtained by theoretical equations. If Fload1 and Fload2 are determined by the theoretical formula, the engine specifications (Vm, V
e) and the microcomputer specification (Δt) only need to be given, so there is no need to match with the actual machine.
Even if load1 and Fload2 are used, the pulse width Avtp corresponding to the cylinder air amount can be obtained with high accuracy.

In this example, in order to further accurately calculate the cylinder air amount, a delay system for calculating the air amount is configured as shown in FIG. This is a review of the following points, which is the same as before. In the drawings, the air flow meter is abbreviated as AFM (the same applies to FIGS. 4, 10, and 12).

1) The time constant of the low-pass filter for the air flow meter of the control unit is set from 20 ms to 1.5.
ms in order to reduce the degree of ignition noise removal.

2) The weighted average of the air amount after linearization (described later) is eliminated.

FIG. 4 shows a pulse width Avt corresponding to the cylinder air amount.
This is a routine for calculating p. This routine is repeated at regular intervals (for example, every 4 ms). The operation of this example will be described with reference to this flowchart.

In step 1, the signal from the air flow meter 18 is A / D converted, and in step 2, a constant process (referred to as linearization process) is performed so that the A / D conversion value Qa is proportional to the amount of air. The raised value is stored in a memory called Qs. FIG. 5 shows table characteristics for linearizing.

In step 3, the basic injection pulse width Tp [ms] is calculated from the linearized air amount Qs [g / s] and the engine speed Ne by the following equation.

Tp = (Qs / Ne) · KCONST # (28) In the equation (28), Qs / Ne is an air amount per unit rotation, and this value is multiplied by an air amount-pulse width conversion constant KCONST #. Since the value is Tp, Tp is a signal corresponding to the amount of air per unit rotation.

In step 4, the basic injection pulse width Tp is corrected by the following equation.

Tpflat = Tp · Ktrm (29) Ktrm is a coefficient for correcting errors in the air flow characteristics of the air flow meter and the fuel flow characteristics of the injector for each operating condition of the engine, and trm means trimming.
To determine the trimming coefficient Ktrm, the engine speed N
A 16 × 16 map assigned by e and α-N flow rate Qh0 is referred to with interpolation calculation. Note that Qh0 is a known air flow rate of the throttle valve portion at a steady state determined from the throttle valve opening TVO and the rotation speed Ne.

In step 5, the cylinder air amount-equivalent pulse width Avtp (the initial value is a value corresponding to a charging efficiency of 100%) is calculated by the following equation. Avtp = Tpflat · Fload1 + old Avtp · Fload2 (30) where old Avtp: previous value of Avtp

The two weighted average coefficients Fload in equation (30)
1. The calculation of Fload2 will be described with reference to the flowchart of FIG.

The flowchart of FIG. 6 is executed at every calculation time interval Δt independently of the flow of FIG.

In steps 11 to 13, the in-cylinder fresh air ratio η (n) is calculated using the following approximate expression. Although η (n) is originally determined by suction negative pressure / exhaust pressure, it is determined here depending on the engine load equivalent amount and the rotational speed. η (n) = η _b + (1−η _b ) · η _N (31)

Among them, η _b is a fresh air ratio load dependent component (basic efficiency value). Equivalent engine load (Avtp, Qh
0, etc.), the α-N flow rate Qh0 is used. Avtp
The reason why Avtp is not adopted is that Avtp is affected by the atmospheric pressure and the intake air temperature, and because it is ηb for finding a new Avtp, there is only data 4 ms before and there is a delay.
This is because the calculation time is prolonged because it is necessary to devise a method such as repeatedly calculating Fload1 and Fload2 twice.

Since η (n) changes depending on the rotational speed Ne as shown in FIG. 7, the rotational speed correction rate η _N is introduced with η _b as the basic value (the value when Ne = 0). ing. In FIG. 7, A / (A + B) = η _N.

Specifically, at step 11, the α-N flow rate Qh
The table of 0 to η _{b and} the table of engine speeds Ne to η _N are referred to in step 12 with interpolation calculation. 8 and 9 show the contents of the tables of η _b and η _N , respectively. Η (n) is calculated by substituting the referred η _b and η _N into equation (31) (step 13).

The engine load equivalent (Avtp or Qh
Η (n) can also be determined by directly referring to a map of η with the parameters of 0) and the engine speed Ne.

Using the in-cylinder fresh air ratio η (n) thus obtained and the rotational speed Ne at that time, step 1 is performed.
In step 4, the coefficient Kfl is calculated by the above equation (22). In step 15, η (n) and η (n-1) (η (n)
(Previous value) is used to calculate the change rate β of the in-cylinder fresh air ratio by the above equation (24). In steps 16 and 17, Kfl and β are substituted into equations (26) and (27) obtained using the manifold filling model and the cylinder suction model to calculate two weighted average coefficients Fload1 and Fload2.

In this case, two load average coefficients Fload
1. When preparing Fload2,
Specify the engine specifications (Vm, Ve) of the applicable model in advance.
It only needs to be stored in the OM. In other words, F
In order to create the maps of load1 and Fload2, it is not necessary to match with the actual machine, and the number of steps for matching can be reduced. Even if the applicable model is different, it is only necessary to change the engine specifications accordingly.

Further, since equations (26) and (27) are theoretical equations, in calculating these equations, the actual cylinder air amount should be calculated as long as the delay occurring in the delay system assumed as a premise is small. It can be accurately predicted.

Finally, at step 18, η is calculated for the next calculation.
The value of the memory (n) is transferred to the memory of η (n-1), and the current routine ends.

FIG. 11 shows the calculation results of the two weighted average coefficients Fload1 and Fload2 by the conventional apparatus under the same acceleration conditions and the calculation results of the two weighted average coefficients Fload1 and Fload2 according to the present invention. As shown in the upper part of FIG. 11, according to the present invention, when the in-cylinder fresh air ratio η (n) does not change with time (that is, in each steady state before and after acceleration), β = 1
Fload1 + Fload2 = 1, and when there is a temporal change in η (n) (that is, during acceleration), Fload1 + Fload2
load2 ≠ 1. On the other hand, in the conventional apparatus, F
load1 + Fload2 = 1.

FIG. 12 shows the response waveforms of the air amount during acceleration of the present invention and the conventional device in an overlapping manner. From the figure,
In the present invention as well, the rise delay immediately after acceleration (the fall delay in the initial stage of deceleration during deceleration) and the overshoot in the latter period of acceleration are significantly improved, and the calculation of the weighted average coefficients Fload1 and Fload2 with the conventional device is performed. Due to the difference in the results, the waveform of the present invention is obtained as a waveform closer to the actual change in the air amount in the cylinder than that of the conventional device.

FIG. 13 shows Fload1 and Fload during deceleration.
2 shows the change. Normally, η (t + 1)> η (t) during acceleration, whereas η (t + 1) <during deceleration.
Since it is usually η (t), the result is as shown in FIG.

Now, the two weighted average coefficients are given by (2)
6), instead of the theoretical formula of (27), the calculation can be performed by the following approximate formula. Fload1 = 1 / Kfl (32) Fload2 = (1-Fload1) / β (33)

Here, equation (32) is a theoretical equation (2)
In equation (6), β ≒ 1 and Kfl >> 1 are obtained. Equation (33) is a theoretical equation (27)
It is obtained by transforming the equation as follows. Fload2 = Kfl / (1 + Kfl · β) {(1 + Kfl · β-1) / (1 + Kfl · β)} · (1 / β) ({β ≒ 1) = {1-1 / (1 + Kfl · β)} · (1 / β) = (1-Fload1) / β

There is almost no difference between the calculation result of calculating Avtp by the approximation formulas (32) and (33) thus obtained and the calculation result of Avtp by the theoretical formulas (26) and (27). Has been confirmed by experiments.

Next, FIG. 10 shows the effect of the time constant of the low-pass filter (see FIG. 2) provided for the air flow meter on the lean spike during acceleration. As the time constant increases, the amount of deviation in the lean direction increases. Is getting bigger.

In this case, since the time constant is changed to at least 10 ms in this example, the lean spike during acceleration can be further improved.

In the embodiment, the load signal for referring to the in-cylinder fresh air ratio η (n) is set to Qh0. However, Avtp may be used to repeatedly calculate the weighted average coefficients Fload1, Fload2 and Avtp. In order to cancel the influence of the atmospheric pressure, Avtp / air density (from Tp or the like when the throttle valve is fully opened) may be used as a load signal for referring to η (n). In the embodiment, since the microcomputer performs the digital control, the discrete value system has been described. However, it is needless to say that the microcomputer can be configured as a continuous value system.

Although the case of acceleration is described in FIG. 12, the same applies to the case of deceleration.

[Brief description of the drawings]

FIG. 1 is a system diagram of an embodiment in which an air amount detection device of the present invention is applied to an L-jetronic type fuel injection device.

FIG. 2 is a block diagram of a delay system for calculating an air amount according to the embodiment.

FIG. 3 is a diagram showing a physical model of an intake system.

FIG. 4 is a flowchart for explaining calculation of a pulse width Avtp corresponding to a cylinder air amount.

FIG. 5 is a table characteristic diagram of a linearized air amount Qs.

FIG. 6 shows a cylinder fresh air ratio η (n) and a weighted average coefficient F
It is a flowchart for demonstrating calculation of load1 and Fload2.

FIG. 7 is a characteristic diagram for explaining how to obtain a cylinder fresh air ratio η (n).

FIG. 8 is a table characteristic diagram of a fresh air ratio load dependent component η _b .

FIG. 9 is a table characteristic diagram of a rotation speed correction rate η _N.

FIG. 10 is a characteristic diagram of a lean spike during acceleration.

FIG. 11 shows two weighted average coefficients F during acceleration according to the present invention.
It is a waveform diagram which shows each calculation result of load1 and Fload2 and two weighted average coefficients Fload1 and Fload2 at the time of acceleration by a conventional apparatus.

FIG. 12 is a waveform chart for explaining an operation at the time of acceleration according to the embodiment.

FIG. 13 shows two weighted average coefficients F during deceleration according to the present invention.
It is a waveform diagram which shows each calculation result of load1 and Fload2.

FIG. 14 is a diagram corresponding to the claims of the first invention.

FIG. 15 is a diagram corresponding to a claim of the sixth invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Intake passage 13 Injector 14 Cylinder 17 Intake throttle valve 18 Hot wire air flow meter 21 Crank angle sensor 25 Control unit

Claims (6)

[Claims] An air flow meter that outputs an amount of air in accordance with the amount of air upstream of an intake throttle valve; a means for calculating a signal corresponding to an amount of intake air per unit rotation based on the output of the air flow meter; Means for calculating a second weighted average coefficient in consideration of a temporal change in the intake volume in the cylinder; and a first signal corresponding to the amount of intake air per unit rotation.
Means for calculating the sum of the value obtained by multiplying the weighted average coefficient of the above and the signal obtained by multiplying the signal corresponding to the previous cylinder air amount by the second weighted average coefficient as the signal corresponding to the current cylinder air amount. Characteristic engine air amount detection device. 2. The air amount detecting device for an engine according to claim 1, wherein said two weighted average coefficients are calculated based on a change rate of a fresh air ratio in the cylinder. 3. The change rate of the in-cylinder fresh air ratio is β = η (n−1) / η (n), where β: the change rate of the in-cylinder fresh air ratio, η (n): the in-cylinder fresh air ratio The air amount detection device for an engine according to claim 2, wherein the calculation is performed by the following equation:? 4. The two weighted average coefficients are: Fload1 = 1 / (1 + Kfl · β) Fload2 = Kfl / (1 + Kfl · β) Kfl = (120 · Vm) / (Ve · η (n) · Ne · Δ
t) β = η (n−1) / η (n) where Fload1: first weighted average coefficient Fload2: second weighted average coefficient Kfl: coefficient β: rate of change of fresh air ratio in cylinder Vm: manifold volume Ve: Displacement η (n): Ratio of fresh air in cylinder η (n-1): Previous value of ratio of fresh air in cylinder Ne: Engine speed Δt: Calculation time interval Item 2. An air amount detection device for an engine according to Item 1. 5. The two weighted average coefficients are: Fload1 = 1 / Kfl Fload2 = (1-Fload1) / β Kfl = (120 · Vm) / (Ve · η (n) · Ne · Δ
t) β = η (n−1) / η (n) where Fload1: first weighted average coefficient Fload2: second weighted average coefficient Kfl: coefficient β: rate of change of fresh air ratio in cylinder Vm: manifold volume Ve: Displacement η (n): Ratio of fresh air in cylinder η (n-1): Previous value of ratio of fresh air in cylinder Ne: Engine speed Δt: Calculation time interval Item 2. An air amount detection device for an engine according to Item 1. 6. The low-pass filter 6 connected between the air flow meter and the intake air amount signal calculating means has a time constant of at least within 10 ms. An air amount detection device for an engine as described in the above.