JP2002054950A - Abnormality detection device for three-phase type phase signal generator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an abnormality detection device for a three-phase type phase signal generator hardly receiving an adverse effect of intruding noise on a measurement system and capable of surely detecting the abnormality such as disconnection and a short circuit of a wiring system. SOLUTION: A phase signal S is three-bit information comprising bits u, v and w, and a subroutine YY realizing a temporality determination part is composed of a substandard value (S=0, 7) detection means, determination suspension means y10-y20, a temporality determination means y20, suspension continuation means y45-y75, normal state restoration means y65-y85, intermittent abnormality number accumulation means (variable K) and continuous abnormality number accumulation means (variable J). In y50 and below, abnormal processing is respectively carried out when the substandard values are detected (a) times continuously during a suspension continuation period or (b) times during the suspension continuation period. The suspension of the temporality determination is continued since the abnormality (S=0, 7) is detected until all the inverting events of the respective bits of a value Y10 within the standard thereafter detected for the first time are detected, and the measurement system is restored to its normal state by those operations when the abnormality is recoverable.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an abnormality for detecting an abnormality in an output signal of a three-phase type phase signal generator for outputting a pulse train electric signal composed of three rows of periodic ON / OFF signals having mutually shifted phases. The detection device can be applied to a pulse encoder or the like. More generally, the present invention is applied to a measuring instrument or a control device such as a rotation angle sensor or a linear scale that measures an angle, a length, a position, a speed, and a rotation angular speed using a periodic phase signal, a motor control device, or the like. Can be used.

[0002]

2. Description of the Related Art As a device for detecting a position (or rotation angle), a displacement direction, a displacement speed, and the like of an object by a pulse encoder, for example, Japanese Patent Laid-Open Publication No. Hei 3-4768
9: Steering angle detecting device ”is generally known.

[0003] A two-phase pulse encoder is used in the steering angle detecting device of the above-mentioned patent publication, and an example (cross-sectional view) of a conventional three-phase pulse encoder similar to this is used.
Is illustrated in FIG. The encoder 10 is used to detect the rotation angle θ of the rotating plate 21 attached to the steering shaft 20, similarly to the above-described steering angle detecting device, and includes a slit provided on the rotating plate 21. 22, U phase detection sensor 23, V phase detection sensor 24,
It comprises a W-phase detection sensor 25 and the like. The encoders 10 communicate with each other through a waveform shaping circuit (not shown).
Periodic on / off signals (1 bit information) u, v, w shifted by three periods (2α) are output in parallel (FIG. 12).

FIG. 12 shows an output waveform diagram (a) of the three-phase pulse encoder 10 and a definition table (b) of the phase signal S directly related thereto. Here, the phase signal S is 3-bit information (binary number) defined by the following equation (1).

S = 4u + 2v + w (1)

[0005] For example, when the encoder (rotation angle sensor) is configured as described above, the phase signal S is generated by a factor such as disconnection or short circuit of the wiring system or noise entering from the wiring system.
The value (0, 7) out of the standard may be continuously or temporarily indicated.

[0006]

In the case where the abnormality detected by the detection of such an out-of-specification value is a transient abnormality due to a temporary factor such as noise, etc., depending on the condition of occurrence of the abnormality, In some cases, it is possible to return the encoder (rotation angle sensor) to a normal state.

However, in the conventional rotation angle sensor or the like, whether these abnormalities are transient abnormalities due to temporary factors such as noise, or continuous or permanent abnormalities due to disconnection, short circuit, etc. Because there is no means for determining whether the rotation angle sensor can be returned to a normal state, abnormal measures are taken or abnormal measures are delayed due to a delay in abnormality detection. There is a possibility that the abnormal state may not be detected.

In particular, when the phase signal changes rapidly (fast), the above-mentioned nonstandard value (S = 0, 7) is intermittently detected. It is difficult to detect. For this reason, it is difficult to judge the above-mentioned transient state, and it is difficult to determine whether a transient abnormal state occurs simply by performing a simple division with a predetermined time constant as a boundary value using the abnormal state duration time as one index. In some cases, it is very difficult to determine whether the abnormality is abnormal.

In addition, the above wiring system is physically doubled,
By comparing input signals from both of them, there may be a method of specifying the location where an abnormality has occurred or the type of a cause of a failure. However, depending on such a method, it is possible to eliminate the need for a redundant wiring system. As compared with the method of testing (determining) an abnormality by itself, problems such as a complicated wiring system, a high cost of the wiring system, and a complicated determination process arise.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to prevent an adverse effect of intrusion noise on a measurement system and to detect an abnormality such as disconnection or short circuit of a wiring system. An object of the present invention is to realize a reliable three-phase phase signal generator abnormality detection device.

[0011]

In order to solve the above-mentioned problems, the following means are effective. That is, the first means is an abnormality detection device for detecting an abnormality of an output signal of a three-phase type phase signal generator that outputs a pulse train electric signal composed of three rows of periodic on / off signals having mutually shifted phases. ,
A determination holding means for temporarily holding the determination regarding the once detected abnormality until after the next sampling time, and whether the abnormality is a transient abnormality due to a temporary factor such as noise, or a disconnection, short circuit, etc. And a transient determination means for determining whether the abnormality is a persistent or permanent abnormality.

[0012] The second means is a means for judging the suspension of the first means, wherein the number of occurrences or the number of detections of the intermittently occurring abnormalities are added up during a specific period and accumulated, thereby producing an output signal. Is to provide a hold continuation unit that continues to hold without canceling the hold even if it once indicates a normal value or a value within the standard.

[0013] The third means may be the first or the second means.
In the means, when the reversal of the periodic ON / OFF signal is detected in all three columns after the suspension is started, the suspension is released, and the above abnormality is determined to be a transient abnormality. And a normal state return means for returning to a normal state.

Further, the fourth means includes the first to third means.
In any one of the means, at least from the start of the hold to the time when all the inversion events of the periodic on / off signal are detected in all three columns, the number of intermittent occurrences of the abnormality Alternatively, an intermittent abnormality number accumulating means for accumulating the total number of times of detection is provided.

Further, the fifth means includes the first to fourth means.
In the determination suspending means of any one of the means, when the output signal indicates a normal value or a value within the standard, a continuous abnormality that clears the accumulated value of the number of occurrences or the number of detections of the abnormality that has occurred continuously until then to zero. This is to provide a number-of-times accumulating means.

[0016] The sixth means may include the first to fifth items.
In any one of the means, when the abnormality is a transient abnormality, the output signal compensating means for compensating for a pulse train electric signal, a phase, a serial signal obtained based on the phase, or a related value thereof. It is to prepare.

[0017] The seventh means is an abnormality for detecting an abnormality of an output signal of a three-phase type phase signal generator which outputs a pulse train electric signal composed of three rows of periodic on / off signals having mutually shifted phases. The detection device is provided with an output signal compensating means for compensating for a pulse train electric signal, a phase, a serial signal obtained based on the phase, or a related value thereof when the detected abnormality can be recovered.

The eighth means may be the sixth or seventh means described above.
In the output signal compensating means of the means, the sampling data of the pulse train electric signal, the amount of phase change per unit time,
The above-described compensation is performed based on accumulated data up to the current time, such as a phase change direction or a related value related to these physical quantities.

[0019] The ninth means may be any of the above-mentioned sixth to eighth aspects.
In any one of the means, in connection with the above compensation,
Another feature of the invention is to provide a storage data correction unit for correcting the storage data.

According to a tenth aspect, in a rotation angle sensor having a three-phase type phase signal generator, the abnormality detecting device of any one of the first to ninth means is provided.

According to an eleventh aspect, in the linear scale having the three-phase type phase signal generator, the abnormality detecting device of any one of the first to ninth means is provided.

A twelfth means is a motor control device provided with a rotation angle sensor having a three-phase type phase signal generator, wherein the abnormality control device is provided with any one of the first to ninth means. That is. That is, the motor control device includes the rotation angle sensor of the tenth means.

According to a thirteenth aspect, a power steering system mounted on a vehicle includes the rotation angle sensor according to the tenth aspect or the motor control device according to the twelfth aspect. With the above means, the above-mentioned problem can be solved.

[0024]

Operation and effect of the present invention If the transient judgment means of the present invention is used, the cause of the detected out-of-specification value (cause of abnormality)
Is transient or persistent (or permanent)
Because it is possible to determine whether the rotation angle sensor is normal, when it is possible to return the rotation angle sensor to the normal state,
This eliminates the possibility that the abnormal measures are performed, the abnormal measures are delayed due to the delay of the abnormal detection, or the abnormality cannot be detected sufficiently reliably.

Normally, a temporary abnormality of the output signal within about several tens of milliseconds is often a transient abnormality due to noise intruding from the wiring system, and such a transient abnormality is detected. In such a case, the measurement system can often be returned to a normal state. Under such circumstances, when performing the above-mentioned transient determination, a certain amount of time is required. The determination suspending means of the present invention is for suspending the above-mentioned transient determination during the “time period of a certain amount or more”.

In particular, when the phase signal changes drastically (fast), for example, the abnormal state continuation time is used as one index, and a simple division (determination) is performed using a predetermined time constant as a boundary value.
May be difficult to determine the above-mentioned transient, but the use of the hold continuation means of the present invention makes it possible to detect the number of abnormalities such as non-standard values detected intermittently within a predetermined period. Can be accumulated, so that even if such an abnormality is intermittently detected, the above-described transient determination can be reliably performed.

[0027] Further, for example, after the abnormality (out-of-specification value) is detected, each of the above-mentioned predetermined periods is set to one of the in-specification values (one of S = 1 to 6) detected thereafter. This is a period until all the bit inversion events are detected. If the predetermined period is set in this way, it is determined whether the once detected abnormality is a transient abnormality based on the number of non-standard values detected intermittently or continuously during this period. be able to.

This determination is made when a continuous or permanent abnormality has occurred in at least one of the wires u, v, and w due to disconnection, short circuit, etc., and then each bit (phase signal S
) Are based on the fact that not all of the inversion events of each of these bits are detected, and if all of the inversion events of each of these bits are respectively detected, the abnormality detected before that ( (Out-of-specification value) can be determined as a transient abnormality due to a temporary factor such as noise.

Therefore, for example, if the above-mentioned normal state restoring means is also used at the same time, in such a case, the above-mentioned suspension is released, the above-mentioned abnormality is judged as a transient abnormality, and the normal state is restored. can do.

The intermittent abnormality count accumulating means of the present invention is used when the above-mentioned abnormality (non-standard value or the like) is intermittently detected. This is used when non-standard values are detected continuously. With these accumulating means, the above-mentioned transient judgment can be performed regardless of whether an abnormality (such as a nonstandard value) is detected intermittently or continuously.

Therefore, for example, if the above-mentioned holding continuation means is used at the same time, the phase signal S changes greatly (fast).
Even in such cases, by detecting and counting the abnormalities (non-standard values and the like) intermittently, it is possible to continue the determination while suspending the determination of the abnormalities.

When the system (measurement system) is returned to the normal state from the temporary abnormal state, the update to the correct value to be performed in each sampling cycle is temporarily suspended due to disturbance such as noise. It is necessary to update (compensate) the value of the position or rotation angle or the like that has been set to a correct value.
The output signal compensating means of the present invention is extremely useful, for example, when such an abnormal state is recovered (when a non-standard value is detected next time after a non-standard value is detected) or when the hold of a determination is released.

In particular, when the phase signal S changes greatly (fast), the updating (compensation) may not be easy. However, according to the present invention, such updating of the position or rotation angle or the like is not possible. Compensation can be accurately performed based on past accumulated data.

While the judgment is pending, there is a case where these stored data cannot be updated accurately. For example, a case where an out-of-specification value is detected corresponds to this. However, even in such a case, according to the stored data correction means of the present invention, the correction processing of these past stored data is performed at any time, such as at the time of recovery from an abnormal state, or at the time of release of the hold of the determination, and the like. Also, even when entering the determination suspension state, the output signal compensating means can be operated with high accuracy.

The invention relating to the abnormality detecting device of the three-phase type phase signal generator relates to, for example, a motor control device or a steering angle sensor (for a brushless motor) mounted on a vehicle-mounted power steering device (power steering system). Steering sensor). More generally, the present invention can be applied not only to the rotation angle sensor but also to a linear scale or the like that measures the length and position using a three-phase type phase signal generator.

[0036]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described based on specific embodiments. However, the present invention is not limited to the embodiments described below. FIG. 1 is a logical configuration diagram illustrating a logical main configuration of the abnormality detection device of the three-phase phase signal generator according to the present embodiment.

The three-phase type phase signal generator 100 corresponds to, for example, the pulse detection sensor (U-phase detection sensor 23, V-phase detection sensor 24, W-phase detection sensor 25) shown in FIG. Physical quantities U, V, W (eg, light intensity)
Are converted by the waveform shaping circuit 110 into on / off signals u, v, w (one-bit information) shown in FIG. 12 and input to the computer. Hereinafter, the operation and functions of a computer that receives the on / off signals u, v, and w and executes predetermined information processing will be described.

The abnormality detecting section 200 receives the serial signal (θ)
The calculation unit 210, the transient determination unit 220, and a control program (a main program P to be described later) that controls these units.
P). The serial signal calculating section 210 is a program embodied by a subroutine XX described later, and outputs a serial signal θ (rotation angle) calculated using an output signal compensating means described later to the application program 400.

The transient judgment section 220 is a program embodied by a subroutine YY to be described later, and a judgment holding section 221 for temporarily holding the judgment regarding the once detected abnormality until the next sampling time. Whether the abnormality is a transient abnormality due to temporary factors such as noise,
Alternatively, there is provided a transient determination means for determining whether the abnormality is persistent or permanent due to disconnection, short circuit, or the like.

When the serial signal calculation section 210 or the transient determination section 220 detects an abnormality that cannot be restored (recovered) to normal, the abnormality processing section 250 is started. The abnormality processing unit 250 may be configured in association with the application program 400 described above.

FIG. 2 is a general flowchart illustrating an outline of a processing procedure of the main program PP for controlling the operation of the abnormality detecting device according to the present embodiment. In the program PP, first, the value θ ₀ of the reference point of the rotation angle (serial signal) is determined in step p05. The value θ _{0 of the} reference point may be periodically updated (corrected), for example, asynchronously with the program PP.

Next, at step p10, the time variable T is set to be smaller than the current time (timer time) t by the on / off signal u,
The value is set to a value larger by the sampling interval ΔT of v and w. At Step p15, the abnormality flag F1 is initialized to 0. In step p20, on / off signals u, v, w
From the waveform shaping circuit 110.

At step p25, an initial value (previous value) S0 of the phase signal S defined by the above equation (1) is set (initialized). In step p30, the phase signals S, S
An initial value (previous value) d0 of the phase displacement amount d at one sampling interval ΔT that can be determined from 0 is set (initialized). This displacement d is, for example, as shown in FIG.
In (a), when the phase signal S changes from “4” to “6” immediately on the right, it is a numerical value such that d = + 1, and this initial value may be temporarily set to 0.

At step p35, the value of the rotation angle θ is initialized according to the following equation (2).

Θ = θ ₀ + αd0 (2) where α is the quantity shown in FIG.
This is a quantity indicating the width of the angle region occupied by one phase signal S. That is, for example, when d = + 1 as described above, θ increases by one phase (α).
For example, in the case of a steering sensor (steering angle sensor) as shown in FIG. 11, the shorter the set interval (set cycle) of the slit 22, the smaller the value of α becomes.

At step p40, other variables are initialized. Here, the variables to be initialized include, for example,
There are variables J, K, and Y0 used in the subroutine YY, and a variable X used in the subroutine XX. The variables J, K,
The initial values of Y0 and X are each 0.

At step p45, the process waits until t ≧ T. Here, t is the timer time (current time) described above. In step p50, the on / off signal u,
v and w are input from the waveform shaping circuit 110. At Step p55, the value of the phase signal S defined by the above equation (1) is updated.

In step p60, a subroutine YY (FIG. 3) for realizing the above-mentioned transient judgment section 220 is called and executed. In step p65, a subroutine XX (FIG. 4) for realizing the serial signal calculating section 210 is called and executed. At step p70, the abnormality flag F
The value of 1 is checked. If the value is 0, the process proceeds to step p75; otherwise, the process proceeds to step p90.

At step p75, the value of the serial signal (rotation angle) θ is output. In step p80, the value of the time variable T is set to a value larger than the previously
Is reset, and the control is moved to step p45. In step p90, a predetermined end process is executed to complete the execution of the main program PP.

According to the above control program (main program PP), each processing program (subroutine XX, YY) of the serial signal calculation section 210 and the transient determination section 220 constituting the abnormality detection section 200 (FIG. 1) is executed. Can be controlled.

FIG. 3 is a flowchart exemplifying a processing procedure of a subroutine YY which is called and executed by the main program PP and realizes a transient determination section (transient determining means). The subroutine YY implements each of the following units out of the units of the present invention. (1) Out-of-specification value detecting means: Step y05 (2) Judgment suspension means: Steps y10 to y20 (3) Transient judgment means: Step y20 (4) Suspension continuation means: Steps y45 to y75 (5) Return to normal state Means: Steps y65 to y85 (6) Intermittent abnormality count accumulating means: Variable K (Step y1
(5, step y80) (7) Means for accumulating continuous abnormal times: variable J (step y1
0, step y40)

In the subroutine YY, first, the value of the phase signal S is checked in step y05, and if the value is out of the standard (0 or 7), step y10
Otherwise, the process proceeds to step y40. At step y10, the value of the variable J is increased by one. At step y15, the value of the variable K is increased by one.

In step y20, the values of the variables J and K are checked, and if at least one of the following equations (3) and (4) is satisfied, the process proceeds to step y25; Caller (main program PP)
Return the process to

(3) J ≧ a (3)

[Expression 4] K ≧ b (4)

Here, a and b are predetermined constants. According to such a determination (step y20), the nonstandard value (S = 0 or 7) is detected by the nonstandard value detecting means (step y0).
When a is detected a times consecutively in 5), an abnormal process (step y25) is executed, and a non-standard value (S = 0 or S = 0) is set during the hold continuation period determined by the hold continuation means (steps y45 to y75). Similarly, when b) is intermittently detected b times, the abnormality processing (step y25) is executed. As the values of the constants a and b, optimal values may be selected based on the duration of noise that can be assumed, the configuration and use of the measurement system, and the like.

At step y25, abnormal processing corresponding to the occurrence of disconnection or short circuit of the wiring system is executed. Step y3
At 0, the value of the abnormality flag F1 is set to 1.

At step y40, the value of the variable J is cleared to zero. As a result, the number of times abnormalities (non-standard values) are continuously detected is counted in step y10 (continuous abnormal number accumulating means). In step y45,
The value of the variable K is checked. If the value is 0, the process is returned to the calling source (main program PP).
Transfer processing to As a result, the following processing is executed during the above-mentioned suspension continuation period.

That is, at step y50, the value of the variable Y0 is confirmed. If the value is 0, the process proceeds to step y55; otherwise, the process proceeds to step y65. As a result, the following initial settings (step y55, step y55) are performed only when K = 1 (that is, after the above-described suspension continuation period is started) and then step y50 is first executed.
60) is executed. This initial setting is for defining the above-mentioned suspension continuation period.

That is, in step y55, the current value of the phase signal S is stored in the variable Y0. Thereby, after the abnormality (non-standard value) is detected in step y05, the value of the phase signal S within the standard (one of 1, 2, 3, 4, 5, and 6) detected first after that is detected. Is stored in Y0. Also,
At step y60, the value of the variable Y2 is initialized to 0.

At step y65, the variable Y0 and the current phase signal S
Is obtained for each corresponding bit, and the result is stored in a variable Y1. By this processing, the inverted bit of the current phase signal S for each bit of the variable Y0 is obtained every time. Therefore, Y1 is always initialized to 0 by this step y65 only immediately after the above-mentioned initialization (step y55, step y60) is executed.

At step y70, the inverted bits obtained at step y65 are cumulatively stored in variable Y2 for each corresponding bit. That is, after the value of Y0 is determined by this process, if there is a bit (u bit, v bit, or w bit) inverted in the phase signal S, that bit is set to the corresponding bit of the variable Y2. Remembered,
The value of each corresponding bit of the variable Y2 that has once become "1" is held until the next time step y60 (or step p90 in FIG. 2) is executed.

In step y75, the value of the variable Y2 is confirmed. If 7 (all bits have been inverted), the flow advances to step y80.
Otherwise, the processing is shifted to the calling source (main program PP). As a result, the above-mentioned holding is continued (holding continuation period) from the detection of the abnormality (non-standard value) to the detection of all the inversion events of each bit of the phase signal S (holding continuation period). means). However, since the step y55 is executed only when the phase signal S is within the standard value, the transition between the two states of all bits 1 (S = 7) and all bits 0 (S = 0) is performed by step y75. Are not detected as an all-bit inversion event.

At step y80, the value of the variable K is returned to 0. In step y85, the value of the variable Y0 is returned to 0. That is, when the value of the variable Y2 is 7 (all bits have been inverted), the suspension of the determination is canceled in steps y80 and y85, and the state is returned to the normal state (normal state returning means).

By the operation (action) of these programs, it is possible to determine whether or not the abnormality once detected is transient, and if the abnormality is a recoverable abnormality, Based on this determination, it is possible to return the measurement system to a normal state.

FIG. 4 is a flowchart illustrating a processing procedure of the serial signal calculation unit (subroutine XX) called and executed by the main program PP. The main functions of this program (subroutine XX) are as follows:
One. (1) A serial signal (θ) such as a rotation angle is calculated based on the input phase signal S. (2) When an abnormality is found in the displacement speed of the phase signal S, an abnormal process for abnormally terminating the main measurement system that outputs the serial signal (θ) is executed.

The following means are realized by the subroutine XX (FIG. 4). These means are for accurately obtaining the serial signal (θ) to be output, and are compensated by these means at the time of recovery from an abnormal state (the next time a non-standard value is detected after a non-standard value is detected). The output serial signal (θ) can be output. (A) Output signal compensating means: step x30 (FIG. 10: xs240, xs260) (b) Stored data storage means: step x15 (FIG. 5: xs040), step x25 (FIG. 6: xs140), step x30 (FIG. 10: xs270), step x50 (c) Accumulated data correction means: step x30 (FIG. 10: xs270)

In the subroutine XX, first, in step x05, each subroutine (XSUB1, XS
Return code R updated when abnormality is detected in UB2)
Initialize C to 0. In step x10, the current phase signal S is determined.
If 6), the process proceeds to Step x20.

At step x15, a subroutine XSUB0 (FIG. 5) to be described later is called and executed. In step x20, the phase signal of the previous time (one sampling cycle ΔT before) stored in the variable S0 is determined. If the value is out of specification (0, 7), the process proceeds to step x30. Move the process to At step x25, a subroutine XSUB1 (FIG. 6), which will be described in detail later, is called and executed. In step x30, a subroutine XSUB2 (FIG. 10) described later is called and executed.

Subroutines XSUB0, XSUB1, X
After executing any one of SUB2, the process proceeds to step x40. At step x40, the return code RC is determined. If the return code RC is 0, the process proceeds to step 45; otherwise, the process proceeds to step x60. Step 45
Then, the value of the current serial signal θ is obtained by adding αd to the previous θ. Here, d is one of the above subroutines (XSUB0, XSUB1, XSUB2)
), And indicates the amount of displacement (forward or backward) of the phase specified by the phase signal S. Further, α is a constant represented by the equation (2).

At step x50, the stored data {θ} immediately before θ is updated. The stored data {θ} can be used for estimating the phase displacement speed or the like in the subroutine XSUB2 (FIG. 10) called at step x30. The process of updating the accumulated data {θ} in this step x50 may be executed based on the accumulated data {d} in the most recent past of d updated this time in the subroutine XX,
Alternatively, it may be executed based on the value of θ updated this time in the subroutine XX (step x45). For example,
Immediately after execution of step x30, the stored data {θ} is updated by the former method.

At step x55, the current value of the phase signal S is stored in the variable S0. At step x60, an abnormal process is executed. This abnormality processing is performed in each subroutine (XS
UB1 and XSUB2) may be configured to execute the corresponding abnormal processing in accordance with the value of the return code RC set. At step x60, the value of the abnormality flag F1 is set to 2. With the above processing, the accurate value of the serial signal (θ) is updated every time.

FIG. 5 is a flowchart illustrating the processing procedure of subroutine XSUB0 called and executed from subroutine XX. In this program, first, in step xs010, 0 is set in the main program PP.
Is checked, and if the value is 0, the process proceeds to step xs020. Otherwise, the process proceeds to step xs020.
Move the process to s030. In step xs020, the value S0 of the previous phase signal is saved in the variable X. The variable X is a variable referred to in the subroutine XSUB2 (FIG. 10). By the operation of these two steps xs010 and xs020, the last detected standard value (1, 2, 3, 4,
(Any one of 5 and 6) is stored in the variable X.

At step xs030, the phase displacement d is set to zero. This is tentatively set as a temporary value because the current phase signal S is a non-standard value (0 or 7) and the displacement amount d of the current phase cannot be specified. In step xs040, the accumulated data {d} of the displacement amount d in the most recent past phase is updated. For example, if the accumulated data {d} is composed of I pieces of numerical data (the numerical value of the displacement amount d), the I-th data (the oldest data) that goes back up in the past is deleted, and at step xs030 The updated data (displacement amount d = 0) is registered as the latest data in the accumulated data {d}.

FIG. 6 is a flowchart illustrating a processing procedure of subroutine XSUB1 called and executed from subroutine XX. This subroutine XSUB1
Is executed when the values S0 and S of the previous and current phase signals are both within standard values. In this program,
First, in step xs110, the phase displacement amount d is determined using the function D (S0, S) shown in FIG. Here, FIG. 7 shows a function D for obtaining the phase displacement d from the phase signal S and the previous phase signal S0.
9 shows a definition table for defining (S0, S). For example, when S0 = 4 and S = 6, d = + 1.

Next, at step xs120, the value of d is confirmed. If d = 3, the process proceeds to step xs150, otherwise to xs130. This process
Based on the physical precondition of the present measurement system that the phase (phase signal S) does not change more than 1/4 cycle during one sampling interval (ΔT), the abnormality of the phase change speed is determined. It is for detecting.

In step xs130, the value d0 of the displacement amount of the previous phase (the value of the latest entry of {d})
And the sum (| d0 | + | d |) of the absolute values of the values of the displacement amount d of the current phase is checked. If the value is 4, the process proceeds to step xs150.
The processing moves to 140. This processing is based on the physical precondition of the present measurement system that the phase (phase signal S) does not change more than 1/4 cycle during one sampling interval (ΔT). This is for detecting a speed abnormality. Also, by this determination, “d0 = + 2, d
= −2 ”, it is possible to detect an abnormality when the phase displacement speed is suddenly reversed.

When these abnormalities are detected, it is determined that the speed is abnormal, and step xs15
At 0, the value of the return code RC is set to 1. Also,
In step xs140, the displacement d obtained in step xs110 is converted to the value of step xs in subroutine XSUB0.
Similarly to 040, the stored data {d} is registered as the latest data.

By the above-described processing, when the values S0 and S of the previous phase signal and the current phase signal are both within the standard values, the updating processing of the phase displacement d and the accumulated data {d} is executed. .

FIG. 8 is a numerical sequence exemplifying a temporal change in the phase signal S and the phase displacement d when noise enters the w-phase. Here, a case of a constant velocity state in which the phase (phase) is displaced in a constant direction for one-sixth period (α) per sampling period (ΔT) is illustrated.

For example, as shown in this figure, when noise as shown in FIG. 3B enters the w-phase (or W-phase), the phase that should originally change from “S: 4 → 6 → 2” The signal S is
A case may occur in which “S: 5 → 7 → 3” is changed. In this case, the displacement amount d of the phase when the transition “S: 5 → 7 (non-standard value)” is detected is determined by the subroutine X
It is determined to be 0 by SUB0 (FIG. 5). Also, "S:
The displacement amount d of the phase when the transition “7 → 3” is detected
(D _{k in} FIG. 8) is a subroutine XSUB2 (FIG. 10)
Is determined by The values of the displacement amounts d of the other phases in FIG. 8 are determined by the subroutine XSUB1 as shown in the figure.

In the figure, the mark ◎ indicates the normal state return means (FIG. 3: steps y65 to y8) by the subroutine YY.
5) indicates the timing at which the determination is suspended. The period from when the nonstandard value 7 is detected to when the sampling trigger indicated by the mark “◎” comes is the above-described determination suspension period.

FIG. 9 shows the output signal compensating means (FIG. 1).
0: xs240, xs260) is a graph showing a straight line L indicating an expected value of a phase displacement velocity (rotational angular velocity) that can be used, and the values i, j, and k of the N axis in FIG. _i, d _j, coincides with the timing (sampling trigger) to calculate the d _k, simultaneously y _i in FIG. 9, y _j, y _k
Is calculated (sampling opportunity). The meaning of each symbol used in FIGS. 8 and 9 is as follows.

N: the number of samplings up to the present time d _j : the value of d in the j-th sampling cycle (N = j) y: the sum of the displacement amounts d of the phases up to the present time (y = Σd) y _j : the j-th time Y at the sampling period (N = j) _m y _k : candidate value of y _k m: integer value to be determined (selected) L: straight line indicating the expected value of the phase displacement velocity (rotation angular velocity)

As shown in FIGS. 8 and 9, the value of the phase signal when N = k is S = 3. The above y at this time
As a candidate for the value y _k of a plurality of values _{m k} shown in the following equation (5)
y _k is considered.

_M y _k = ₀ y _k +6 m (m is an integer) (5)

As the value of the reference value ₀ y _k, a value previously determined for each in-standard value of the phase signal S by the main program PP may be used, or the most recent in-phase (S = 3 ) _May be used as it is (the value of y _j in FIG. 9), or may be obtained by using the following equation (6).

[6] _{0 y k = y k-1} + D (X, 3) ... (6) However, where D is a function of FIG. The variable X is set in step xs020 of the subroutine XSUB0 (FIG. 5).
In the present case in which the phase signal S transitions as shown in FIG. 8, the in-standard value 5 last detected up to the point of N = k is stored.

By substituting equation (6) into equation (5), the following equations (7) and (8) are obtained.

Equation 7] _{m y k = y k-1} + D (X, 3) + 6m (m is an integer) ... (7)

Equation 8] _{d k = m y k -y k} -1 = D (X, 3) + 6m (m is an integer) ... (8)

FIG. 10 is a flowchart illustrating the processing procedure of subroutine XSUB2 called and executed from subroutine XX.
2 is used to calculate the phase displacement d _k from the previous time (N = k−1) using the above equation (8).

In the program XSUB2, first, in step xs210, the last standard value saved in the variable X is restored to the variable S0 (as the previous phase signal). In step xs220, the value of the variable X is initialized to 0.

In step xs230, the value of the phase displacement d (d _k ) is provisionally determined using the function D, as in step xs110 described above. The process of step xs230 corresponds to the process of provisionally setting the value of d _k to D (X, 3) with m = 0 in the above equation (8). Therefore,
For example, in the case of FIG. 8, the value of the displacement d (d _k ) is set to “D (5, 3) = − 2” by the present step xs230.

At step xs240, the value of m in equation (8) (or equation (5)) is determined. For example, in the case as shown in FIGS. 8 and 9, a straight line L as shown in FIG. 9 is determined, and the coordinate point (N,
_{y) = (k, m y} k) may be selected integer m giving.

Such a straight line L is, for example, a coordinate point (N, y) = (j, y _j ) shown in FIG. 9 or the latest past data such as the above-mentioned stored data {d} or the above-mentioned stored data {θ}. Can be determined using predetermined storage data (accumulated data). For example, in the present case (example) shown in FIG. 9, the coordinate point (N, y) located closest to the straight line L =
_{(K, m} y k) as an integer giving, may be selected to "m = 1".

At step xs250, it is checked whether m has been determined as described above. For example, if the parameter m can be determined as described above, the processing is shifted to step xs260. However, the intersection between the straight line L and the straight line N = k is, for example, the coordinate point (N, y) = (k, ₀ y _k ) and the coordinate point (N, y) =
In the case where it is located near the midpoint of (k, ₁ y _k ),
It is difficult to determine m. Therefore, in such a case, the determination of the parameter m is abandoned, and the process proceeds to step xs280. At step xs280, the return code RC is set to 2 and the caller (subroutine X
The process (control) is returned to X).

In step xs260, 6 m is added to the displacement d provisionally set in step xs230 according to equation (8) to determine the displacement d of the correct phase. In step xs270, correction and update of the accumulated data {d} are executed. The correction and updating process, for example, FIG. 8, in the case or the like of FIG. 9 is performed by the averaging processing of a plurality of displacement d _n such as follows.

D _n = (y _k −y _j ) / (k−j) (n is an integer satisfying “j <n ≦ k”, and integers j and k are shown in FIGS. 8 and 9) (9)

[0092] In this way, the value of the rational number d _n d _{j +} 1 ~
re-registered in the newly stored data {d} as a correction value of d _k. However, the value of the displacement d returned to the calling source (subroutine XX) includes the displacement d obtained in step xs260.
(Integer) is used as it is. With such a correction,
For the accumulated data {d}, register a numerical value closer to reality,
It can be stored. This correction is useful when the accumulated data {d} is reused, for example, when the straight line L is accurately determined again next time.

As a criterion for determining the integer j as described above, that is, a criterion for determining the target range of the correction (averaging process), for example, the value of the current (N = k) phase signal (S =
A criterion such as going up in the past until the same phase (S = 3) as in 3) appears can be considered. For example, as can be seen from the example of FIG.
This is because it is considered that there is almost no influence of the noise if it goes up to about N = j (S = 3) in the past. Further, the position of the integer j (FIG. 9) as described above may be determined from the latest phase displacement velocity. The phase displacement speed used here may be a rough value as a rough guide. For example, such a rough value can be easily obtained from the accumulated data {d}.

On the contrary, it is desirable that the data used to determine the value of m be N = j (S = 3) or less of the predetermined stored data. In the example of FIG. 8, the signal goes up in the past until the same phase (S = 3) as the current (N = k) phase signal value (S = 3) appears, and before that time (N = j, S = 3). Stored data ((N, y) =
The straight line L is determined using (i, y _i ) and (N, y) = (j, y _j ).

For example, by the operation of the above-described program (subroutines XX, XSUB0, XSUB1, XSUB2), it is possible to continue to output an accurate serial signal (θ) even when a temporary abnormality occurs. Becomes

The number of entries of the stored data {θ} and {d} may be about 8 to 64, more preferably about 16 to 32. If this number is too large, not only does the overhead associated with data processing increase, but also the accuracy of estimating the phase displacement velocity and the like (rotational angular velocity or its related value) decreases due to distant old data. Not desirable. On the other hand, if the number of entries is too small, the accuracy of estimating the phase displacement speed or the like is lowered due to insufficient data amount, which is not desirable. Further, as the accumulated data, the data {y} of the latest past of y ({d) or the like may be stored and updated.

The above-described correction of the stored data may be performed at the time of restoration to the normal state (indicated by 図 in FIG. 8).
For example, when the normal state is restored in the case of FIG. 8, until the same phase (S = 4) appears at that time (N = j−3 = i + 4)
If the averaging process (equation (9)) is performed on the accumulated data of the latest past after going up to the past), the accumulated data {d} can be completely restored (corrected).

The method of estimating the current serial signal (θ) based on the phase displacement velocity or the like (rotational angular velocity or a related value thereof) estimated from the accumulated data {d} or the like also applies to the aforementioned subroutine XSUB0. Can be adopted. By using such means, the estimated value of the serial signal (θ) can be output to the application program 400 (FIG. 1) even when the current phase signal S is a nonstandard value (0 or 7). As a result, the accuracy of the output signal is improved. In particular, when the application program 400 calculates and uses the differential value of the serial signal (θ), it is expected that the accuracy of such an output signal is particularly improved. Further, the processing of estimating the serial signal (θ) at the time of such a speed abnormality or at the time of detecting a nonstandard value can be performed by the application program 400 (FIG. 1).

The present invention relating to the above-described abnormality detection device for a three-phase type phase signal generator is applicable to, for example, a motor control device and a steering angle sensor for a brushless motor mounted on a vehicle-mounted power steering device (power steering system). (Steering sensor) and the like. More generally, the present invention can be applied not only to the rotation angle sensor but also to a linear scale or the like that measures the length and position using a three-phase type phase signal generator.

[Brief description of the drawings]

FIG. 1 is a logical configuration diagram illustrating a logical main configuration of an abnormality detection device of a three-phase phase signal generator according to an embodiment of the present invention.

FIG. 2 is a general flowchart illustrating an outline of a processing procedure of a main program PP for controlling the operation of the abnormality detection device according to the embodiment of the present invention.

FIG. 3 is a flowchart exemplifying a processing procedure of a subroutine YY which is called and executed by a main program PP and implements a transient determination section (transient determining means).

FIG. 4 is a flowchart illustrating a processing procedure of a subroutine XX for realizing a serial signal (θ) calculating unit, which is called and executed by a main program PP.

FIG. 5 is a flowchart illustrating a processing procedure of a subroutine XSUB0 called and executed from a subroutine XX.

FIG. 6 is a flowchart illustrating a processing procedure of a subroutine XSUB1 called and executed from a subroutine XX.

FIG. 7 is a definition table of a function D (S0, S) for obtaining a phase displacement d from a phase signal S and a previous phase signal S0.

FIG. 8 is a numerical sequence illustrating a temporal change of the phase signal S and the phase displacement d when noise enters the w-phase.

FIG. 9 is a graph illustrating a straight line L indicating an expected value of the rotational angular velocity, which can be used by the output signal compensating means.

FIG. 10 is a flowchart illustrating a processing procedure of a subroutine XSUB2 called and executed from the subroutine XX.

FIG. 11 is a sectional view of a three-phase pulse encoder.

12A is an output waveform diagram of a three-phase pulse encoder, and FIG. 12B is a definition table of a phase signal S.

[Explanation of symbols]

Reference Signs List 100 three-phase signal generator 110 waveform shaping circuit 200 abnormality detecting unit (abnormality detecting device) 210 serial signal (θ) calculating unit 220 transient estimating unit (temporary judging means) 221 judgment Reservation means 250… Abnormal processing section PP… Main program for implementing abnormality detecting section 200 XX… Subroutine for implementing serial signal calculating section 210 YY… Subroutine for implementing transient determining section 220 u, v, w… Periodically ON / Off signal S ... phase signal (S = 4u + 2v + w: 3 bit information) S0 ... previous (one sampling cycle earlier) phase signal S d ... phase variation at one sampling interval (integer) d0 ... previous (one sampling cycle) The amount of displacement d {d} in the previous step)... The accumulated data of the most recent past of the amount of displacement d in the phase D. The sum of the current to the function y ... displacement amount d to obtain the displacement amount d (y = Σd) θ ... rotation angle (serial _{signal: θ = θ 0 + αy)} θ 0 ... the value of the reference point of the rotation angle {theta} ... The most recent accumulated data of the rotation angle (serial signal) θ XOR… Exclusive OR operation sign OR… OR operation sign N… Number of samplings up to the present _dj … At the jth sampling cycle (N = j) showing the expected value of the d values y _j ... j-th sampling period value _m y _k ... candidate value m ... determination of y _k of y at (N = j) (selection) should do integer L ... rotational angular velocity Straight line 10: three-phase pulse encoder (three-phase phase signal generator) 20: steering shaft 21: rotary plate 22: slit 23 to 25: pulse detection sensor

Continuation of the front page F term (reference) 2F077 AA01 CC02 NN02 NN27 PP19 QQ07 TT02 2G030 AA01 AF01 AG07 3D033 CA03 CA17 CA21 CA32

Claims (13)

[Claims] 1. A three-row periodic ON / OFF phase shift.
An abnormality detection device for detecting an abnormality of an output signal of a three-phase type phase signal generator that outputs a pulse train electric signal composed of an OFF signal, wherein the determination regarding the once detected abnormality is temporarily performed until after the next sampling time. A determination suspending means for suspending the abnormality, a transient abnormality for determining whether the abnormality is a transient abnormality due to a temporary factor such as noise, or a permanent or permanent abnormality due to disconnection, short circuit, etc. An abnormality detection device for a three-phase phase signal generator, comprising: a determination unit. 2. The determination suspension means according to claim 1, wherein said output signal indicates a normal value or a standard value once by accumulating the number of occurrences or the number of detections of said intermittent abnormality during a specific period. The abnormality detecting device for a three-phase phase signal generator according to claim 1, further comprising a holding continuation unit that continues the holding without canceling the holding. 3. When the reversal of the periodic on / off signal is detected over all three columns after the suspension is started, the suspension is released and the abnormality is temporarily detected. 3. The abnormality detecting device for a three-phase phase signal generator according to claim 1, further comprising a normal state returning means for returning to a normal state upon judging an abnormality. 4. An intermittent occurrence of the abnormality at least from the start of the hold until the inversion events of the periodic on / off signal are all detected over the three columns. 4. The device according to claim 1, further comprising an intermittent abnormality frequency accumulating unit that accumulates the number of times or the number of times of detection. 5.
Abnormality detector for phase type phase signal generator. 5. The determination suspending means, when the output signal indicates a normal value or a value within a standard, continuously resets the accumulated value of the number of occurrences or the number of detections of the abnormality which has been continuously generated up to that time to zero. The abnormality detection device for a three-phase phase signal generator according to any one of claims 1 to 4, further comprising abnormality number accumulation means. 6. An output signal compensating means for compensating the pulse train electric signal, the phase, a serial signal obtained based on the phase, or a related value thereof when the abnormality is a transient abnormality. The abnormality detection device for a three-phase phase signal generator according to any one of claims 1 to 5, further comprising: 7. A method of periodically turning on / off three rows out of phase with each other.
An abnormality detection device that detects an abnormality of an output signal of a three-phase phase signal generator that outputs a pulse train electric signal including an OFF signal, wherein the pulse train electric signal is provided when the detected abnormality can be recovered. An abnormality detection device for a three-phase phase signal generator, comprising: output signal compensating means for compensating for the phase, a serial signal obtained based on the phase, or a related value thereof. 8. The output signal compensating means is based on accumulated data up to the current time, such as sampling data of the pulse train electric signal, a phase change amount per unit time, a phase change direction, or a related value relating to these physical quantities. 8. The abnormality detection device for a three-phase phase signal generator according to claim 6, wherein the compensation is performed by performing the compensation. 9. The three-phase phase signal generator according to claim 6, further comprising a storage data correction unit that corrects the storage data in connection with the compensation. Device abnormality detector. 10. A rotation angle sensor comprising the abnormality detection device for a three-phase phase signal generator according to claim 1. Description: 11. A linear scale comprising the abnormality detection device for a three-phase phase signal generator according to claim 1. Description: 12. A motor control device comprising the rotation angle sensor according to claim 10. 13. A power steering system mounted on a vehicle, comprising the rotation angle sensor according to claim 10 or the motor control device according to claim 12.