JP5299312B2 - Inertial force sensor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an inertia force sensor device performing a comparative diagnosis for two inertia force sensors more accurately than conventional sensor devices. <P>SOLUTION: Each time a detection value of an offset-corrected first angular velocity sensor becomes a prescribed value that can be taken during vehicle turning, the detection frequency for each difference value between entered detection values of two angular velocity sensors is stored, and the difference value with the highest detection frequency is extracted from the stored detection frequency to determine whether or not the difference value is matched with a reference value. When the difference value is not matched with the reference value, if no malfunction such as a sudden change in the characteristics and aging occurs, the reference value is rewritten into an extraction value. The detection value of a second angular velocity sensor after offset correction is corrected based on the reference value, and when a difference value between the corrected value and the detection value of the first angular velocity sensor after offset correction exceeds a given determination threshold, it is determined that either one is faulty. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to an inertial force sensor device that diagnoses a failure of an inertial force sensor by comparing outputs of a pair of inertial force sensors provided in a vehicle with each other.

  Conventionally, as shown in, for example, Patent Document 1, the inertial force detected by the first inertial force sensor and the inertial force detected by the second inertial force sensor that detects the same inertial force as the first inertial force sensor are disclosed. And a vehicle control device (inertial force sensor device) including a failure diagnosis unit that diagnoses a failure based on the comparison result and outputs a diagnosis result is known. Patent Document 1 describes the use of a yaw rate sensor (angular velocity sensor) as two inertial force sensors.

  On the other hand, the output of the angular velocity sensor is ideally zero when the vehicle is stopped, and if it is not zero, an offset is considered to have occurred. Thus, a technique for correcting the offset of the angular velocity sensor based on the output of the angular velocity sensor while the vehicle is stopped is known (for example, see Patent Document 2).

JP 2008-247053 A Japanese Patent No. 3404905

  By the way, the sensor output with respect to the angular velocity of the angular velocity sensor has individual differences depending on the offset and sensitivity. The sensitivity is a change amount (so-called inclination) of the sensor output with respect to the inertial force. Also, variations in offset and sensitivity occur due to variations in the manufacturing process, changes with time (aging), and the like.

  The offset of the angular velocity sensor can be corrected by signal processing based on, for example, the sensor output when the vehicle is stopped as described above.

  On the other hand, the sensitivity of the angular velocity sensor cannot be corrected in the market because there is no method for detecting the true angular velocity (which angular velocity is the detected output during vehicle turning) from the sensor output or the like. For this reason, as shown in FIG. 1, due to the difference in sensitivity between the two angular velocity sensors (the first angular velocity sensor and the second angular velocity sensor), the difference between the outputs of the two angular velocity sensors increases as the angular velocity increases. In FIG. 1, in the two angular velocity sensors, the offsets are set to the same value so that the output difference due to the difference in sensitivity can be easily understood.

  In a configuration in which the difference (difference value) between the outputs of two angular velocity sensors mounted redundantly is determined with a predetermined determination threshold value and a failure diagnosis is performed, the determination threshold value is set with a predetermined margin in consideration of erroneous determination. . Conventionally, for example, in a state where the offset difference is corrected, the difference value of the sensor output is larger as the angular velocity is larger due to the influence of the sensitivity difference as described above. Therefore, the determination threshold is set to the angular velocity (angular velocity sensor as shown in FIG. It was necessary to set so that the larger the output was, the larger the value was.

  For this reason, when an inertial force sensor device having such a determination threshold is applied to, for example, a vehicle skid prevention device, the judgment threshold becomes a large value in a skid state with a high angular velocity that requires accuracy, and the control accuracy for preventing skid is limited. There was a problem that occurred. That is, since the variation in sensor output increases as the angular velocity increases, the sensor output is handled as having some error, and cannot be controlled with high accuracy.

  Regarding sensitivity differences due to variations in the manufacturing process, inspection is performed at the initial stage before use in the market, such as when the sensor is shipped or when the sensor is assembled in a vehicle, and signal processing is performed based on this inspection result. It can also be corrected by. However, as described above, the sensitivity difference between the two angular velocity sensors may change due to changes over time, so that the change is taken into account, and the larger the angular velocity (output of the angular velocity sensor), the larger the value is determined. A threshold must be set.

  In view of the above problems, an object of the present invention is to provide an inertial force sensor device that can compare and diagnose two inertial force sensors with higher accuracy than before.

In order to achieve the above object, an inertial force sensor device according to claim 1 comprises:
A first inertial force sensor and a second inertial force sensor of the same configuration that are provided in a vehicle and output a detection signal corresponding to the inertial force;
The inertial force value corresponding to the detection signals of the two inertial force sensors and the inertial force value after offset correction is inputted, and the inertial force value of the offset-corrected first inertial force sensor is set in advance. A number-of-times storage means for storing the number of detections for each difference value of the inertial force values of the two input inertial force sensors each time when the predetermined value that can be taken when the inertial force acts on
Non-volatile correction value storage means that is rewritable and stores the reference value;
A determination means for extracting a difference value having the largest number of detections from the number storage means and determining whether or not the extracted value matches a reference value;
When the determination means determines that the extracted value does not match the reference value, the extracted value is compared with the reference value, and when the difference between the extracted value and the reference value exceeds a preset first specified value , Characteristic sudden change diagnostic means for determining that there is a failure,
When the determination means determines that the extracted value does not match the reference value, the extracted value is compared with a preset initial fixed value, and a difference between the extracted value and the initial fixed value is set in advance. When the specified value exceeds 2, a time-change diagnostic means for determining a failure,
An update means for rewriting the reference value stored in the correction value storage means to the extracted value when it is determined that there is no failure in the characteristic sudden change diagnosis means and the temporal change diagnosis means;
Based on the reference value stored in the correction value storage means, the deviation due to the sensitivity difference between the inertial force values of the two inertial force sensors after the offset correction is corrected, and the two inertial force sensors subjected to the offset correction and the sensitivity difference correction are corrected. And a comparative diagnosis unit that determines that one of the two is a failure when the difference value of the inertial force value exceeds a predetermined determination threshold value.

  The deviation (difference) between the inertial force values of the two inertial force sensors that have been offset-corrected is due to the sensitivity difference between the two inertial force sensors. Therefore, when the sensitivity difference between the two inertial force sensors is corrected, the inertial force values of the two inertial force sensors coincide with each other when there is no failure.

  In the present invention, as the predetermined value, a value that can be taken by the first inertial force sensor when an inertial force is applied to the vehicle in an offset-corrected state (for example, when the angular velocity sensor is turning) is preset. Yes. The number-of-times storage means stores the number of detections for each differential value of the inertial force values of the two inertial force sensors corrected for offset each time the inertial force value of the offset-corrected first inertial force sensor becomes a predetermined value. Is done.

  The difference value having the largest number of detections (occurrence frequency) extracted by the determination means is the two inertia force sensors corrected by offset when the inertia force value of the first inertia force sensor corrected by offset is a predetermined value. The difference value due to the sensitivity difference. Thus, by extracting the difference value having the highest occurrence frequency, a sensitivity difference described later can be obtained with high accuracy. Then, the determination unit extracts the difference value having the largest number of detections, and compares the extracted difference value (extraction value) with the reference value stored in the correction value storage unit.

  The reference value is difference value data having the highest number of detection times stored in the correction value storage unit at the time of comparison by the determination unit. That is, the determination means compares the latest extracted value with the current extracted value (reference value). Then, when it is determined that the extracted value is suitable for updating the reference value, the updating unit rewrites the latest difference value (extracted value) from which the reference value has been extracted.

  Based on the reference value set in this way, the comparison diagnosis unit corrects the deviation due to the sensitivity difference between the inertial force values of the two inertial force sensors that have been offset-corrected. Thereby, the difference between the inertial force values of the two inertial force sensors after the sensitivity difference correction can be ideally zero.

  Here, with respect to the inertial force value of one inertial force sensor (first inertial force sensor) whose offset has been corrected, the difference value between the inertial force values of the two inertial force sensors whose offset has been corrected is linear (linear function) In other words, the sensitivity difference correction will be described in the case where the inertial force values of the two inertial force sensors change linearly with respect to the acting inertial force.

  When the inertial force value of the first inertial force sensor subjected to the offset correction is zero, the difference value between the inertial force values of the two inertial force sensors subjected to the offset correction is also zero. Further, when the inertial force value of the first inertial force sensor is the predetermined value (for example, α1) set in advance, the difference value between the inertial force values of the two inertial force sensors becomes the reference value (for example, β1). From these two points (0, 0), (α1, β1), the amount of change (inclination) of the inertial force values of the two inertial force sensors after offset correction with respect to the inertial force value of the first inertial force sensor after offset correction. That is, the sensitivity difference becomes clear.

  Accordingly, for example, if the inertial force value of the first inertial force sensor that has been offset-corrected is Y1a and the inertial force value of the second inertial force sensor that has been offset-corrected is Y2a, the inertial force value of the second inertial force sensor that has been offset-corrected. When the sensitivity difference (β1 / α1) · Y1a is subtracted from Y2a, the inertial force value {Y2a− (β1 / α1) · Y1a} of the second inertial force sensor from which the sensitivity difference is subtracted is obtained when there is no failure. This substantially coincides with the inertial force value Y1a of the first inertial force sensor whose offset has been corrected.

  As described above, if the difference value having the largest number of detection times (occurrence frequency) is used for the sensitivity difference correction among the difference values of the inertia force values of the two inertia force sensors subjected to the offset correction, the inertia forces of the two inertia force sensors are used. The value difference value is a substantially constant value regardless of the magnitude of the inertia force when there is no failure such as a short circuit. Therefore, the determination threshold can be set to a substantially constant value regardless of the inertial force. That is, the determination threshold can be reduced even if the inertial force value is large as compared with the conventional case.

  From the above, according to the inertial force sensor device of the present invention, it is possible to make a comparative diagnosis of two inertial force sensors with higher accuracy than in the past.

  In the present invention, when the difference between the extracted value and the reference value stored in the correction value storage means is large, the characteristic sudden change diagnosis means can determine that a failure has occurred due to the characteristic sudden change. On the other hand, when the difference between the initial fixed value set at the time of shipment of the sensor device or the assembly of the vehicle and the extracted value is large, the change over time is accumulated by the time change diagnosis means, resulting in unacceptable fluctuations, resulting in failure. Can be determined.

  As described above, in the present invention, even if the difference value (extracted value) having the largest number of detections is different from the reference value stored in the correction value storage means, the above-mentioned is described before rewriting the reference value to the extracted value. The reference value is updated after determining whether it is appropriate to use the extracted value as a new reference value by the two diagnostic means. Therefore, the reliability of the comparative diagnosis between the two inertial force sensors can be improved.

  In addition, as the predetermined value that can be taken when the inertial force is applied, a difference in sensitivity between the two inertial force sensors clearly appears (the difference between the inertial force values subjected to offset correction is large) and is frequently used when the vehicle is used. It is advisable to set in advance a possible value.

  When both of the two inertial force sensors are angular velocity sensors, as described in claim 2, an angular velocity value corresponding to a detection signal of the two angular velocity sensors is input, and an offset correction unit that corrects an offset of the angular velocity value; A stop determination unit that determines that the vehicle is stopped, and the offset correction unit is configured to set the angular velocity values of the two angular velocity sensors that are input when the stop determination unit determines that the vehicle is stopped. Based on this, it is possible to employ a configuration in which the angular velocity values of the two angular velocity sensors are respectively offset-corrected.

  In this case, the inertial force values of the two inertial force sensors are input to the offset correcting unit and offset correction is performed, and the inertial force values corrected by the offset are input to the number storage unit.

  According to a third aspect of the present invention, the number-of-times storage unit stores the number of detections of the difference value for each of a plurality of predetermined values, and the correction value storage unit stores a reference value for each of the predetermined values. It is also good.

  If there are two predetermined values that can be taken when the inertial force acts on the vehicle, there are three points, including the point where the inertial force does not act (zero output of the first inertial force sensor after offset, zero difference value). The amount of change in the inertial force values of the two inertial force sensors after offset correction with respect to the inertial force value of the first inertial force sensor after offset correction becomes clear by approximating a non-linear line from three points. If the predetermined value is 3 points or more, the amount of change in the inertial force values of the two inertial force sensors becomes clear without using a point where the inertial force does not act. In this way, the sensitivity difference can be corrected even for a non-linear line. As the predetermined value is increased, the sensitivity difference can be corrected more accurately for the non-linear line.

Next, the inertial force sensor device according to claim 4 is:
A first inertial force sensor and a second inertial force sensor of the same configuration that are provided in a vehicle and output a detection signal corresponding to the inertial force;
The difference value of the inertial force value corresponding to the detection signal input from the two inertial force sensors is stored, and each time the difference value is stored, the plurality of stored difference values are used as the inertial value of the first inertial force sensor. An approximate line is calculated by approximation in relation to the force value, and based on this approximate line, a correction value calculating means for obtaining a calculated value of sensitivity difference correction and a calculated value of offset difference correction as a correction calculated value;
Non-volatile correction value storage means that is rewritable and stores sensitivity difference correction values and offset difference correction values as correction values;
A determination unit that determines whether or not the correction calculation value and the corresponding correction value match each other;
When it is determined by the determination means that the correction calculation value does not match the corresponding correction value, if the difference between the correction calculation value and the correction value exceeds a preset first predetermined value, a failure is determined. A characteristic sudden change diagnosis means for determining;
When the determination means determines that the correction calculation value does not match the correction value, the correction calculation value is compared with a preset initial fixed value of the correction, and the difference between the correction calculation value and the initial fixed value is A time-varying diagnostic means for determining a failure when exceeding a corresponding second predetermined value set in advance;
An update means for rewriting the correction value stored in the correction value storage means to the correction calculation value determined as no failure when it is determined that there is no failure in the characteristic sudden change diagnosis means and the temporal change diagnosis means,
Based on the correction value stored in the correction value storage means, the offset difference and sensitivity difference between the inertial force values of the two inertial force sensors are corrected, and the inertial force of the two inertial force sensors subjected to the offset difference correction and the sensitivity difference correction are corrected. Comparing and diagnosing means for determining that one of them is faulty when the difference value of the values exceeds a predetermined determination threshold value.

  In the present invention, the sensitivity difference and the offset difference between the two inertial force sensors are corrected. A correction value calculation means approximates a plurality of stored difference value data to calculate an approximate line, and obtains a calculated value for sensitivity difference correction and a calculated value for offset difference correction as correction calculated values from the approximate line. Can do.

  For example, when the approximate line is a straight line, the intercept of the approximate line corresponds to the offset difference between the inertial force values of the two inertial force sensors, and thus becomes a calculated value for correction. The inclination of the approximated line is the inertial force of the two inertial force sensors. Since this corresponds to the difference in sensitivity of the value, it is a calculated value for sensitivity difference correction.

  Then, the calculated correction calculation values are compared with the correction values stored in the correction value storage means. The correction value is a correction calculation value stored in the correction value storage unit at the time of comparison by the determination unit. That is, the determination unit compares the latest correction calculation value with the current correction calculation value (correction value). When it is determined that the correction calculation value is suitable for updating the correction value, the update means rewrites the correction value to the latest correction calculation value.

  Therefore, in the comparative diagnosis means, the inertial force of the two inertial force sensors is corrected by correcting the sensitivity difference and the offset difference between the inertial force values of the two inertial force sensors based on the correction values (sensitivity correction value and offset correction value). The difference between the values can ideally be zero.

  Here, when the difference between the inertial force values of the two inertial force sensors changes linearly (linear function) with respect to the inertial force value of one inertial force sensor (first inertial force sensor), In this case, sensitivity difference correction will be described in the case where the inertial force values of the two inertial force sensors change linearly with respect to the acting inertial force.

  The above approximate expression is also a linear function (straight line). Assuming that the slope of this approximate expression is e0, the intercept is f0, the inertial force value of the first inertial force sensor is Y1, and the inertial force value of the second inertial force sensor is Y2, it is offset from the inertial force value Y2 of the second inertial force sensor. By subtracting the difference and the sensitivity difference (e0 · Y1 + f0), the inertial force value (Y2− (e0 · Y1 + f0)) of the second inertial force sensor from which the offset difference and the sensitivity difference are subtracted is the case where there is no failure. This substantially coincides with the inertial force value Y1 of the first inertial force sensor.

  As described above, the difference between the inertial force values of the two inertial force sensors corrected for the sensitivity difference and the offset difference based on the correction value is substantially constant regardless of the magnitude of the inertial force when there is no failure such as a short circuit. It becomes the value of. Therefore, the determination threshold can be set to a substantially constant value regardless of the inertial force. That is, the determination threshold can be reduced even if the inertial force value is large as compared with the conventional case.

  From the above, even with the inertial force sensor device of the present invention, the two inertial force sensors can be compared and diagnosed more accurately than in the past.

  According to the present invention, since the offset difference is corrected together with the sensitivity difference, the offset correction of the inertial force value of the inertial force sensor can be made unnecessary in the comparative diagnosis.

  Further, the effects of the sudden characteristic change diagnosis means and the temporal change diagnosis means are the same as those of the first aspect of the invention, so that the description thereof is omitted. As an approximation method, linear approximation (least square method), polynomial approximation, exponential approximation, or the like can be used.

  In the above, as described in claim 5, angular velocity sensors can be employed as the two inertial force sensors.

It is a figure which shows the sensitivity difference of two angular velocity sensors. It is a figure which shows the conventional determination threshold value used for a redundancy comparison. It is a block diagram which shows schematic structure of the angular velocity sensor apparatus which concerns on 1st Embodiment. It is a flowchart which shows the content of the present peak value setting process of the angular velocity sensor apparatus shown in FIG. It is a figure which shows predetermined value (alpha) 1 for adding histogram data. It is the figure which showed the mode of the data addition of a histogram, and the peak value extraction this time. It is a flowchart which shows the content of the comparative diagnosis process of the angular velocity sensor apparatus shown in FIG. It is a figure which shows a sensitivity difference. It is a figure which shows the determination threshold value used for a comparative diagnosis process. It is a figure which shows the modification of the predetermined value for adding histogram data. It is a block diagram which shows schematic structure of the angular velocity sensor apparatus which concerns on 2nd Embodiment. It is a flowchart which shows the content of the present correction value setting process of the angular velocity sensor apparatus shown in FIG. It is a figure which shows an approximate line. It is a flowchart which shows the content of the comparative diagnosis process of the angular velocity sensor apparatus shown in FIG. It is a figure which shows the modification of an approximate line.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
The angular velocity sensor device according to the present embodiment is used for vehicle control, such as a device (vehicle slip prevention device) that maintains the running stability of the vehicle when the posture of the vehicle is disturbed, for example.

  As shown in FIG. 3, the angular velocity sensor device 10 as an inertial force sensor device includes two angular velocity sensors 20 and 21 as inertial force sensors and signal processing for processing the outputs of the angular velocity sensors 20 and 21 as main parts. A microcomputer 30 (hereinafter, referred to as a microcomputer 30) as a unit and an EEPROM 40 as a correction value storage means are provided. In the present embodiment, the microcomputer 30 further processes the output of the wheel speed sensor 50.

  In the angular velocity sensor device 10 according to the present embodiment, the main characteristic part is to correct a deviation due to a difference in sensitivity between the two angular velocity sensors 20 and 21, thereby making a determination used for comparative diagnosis of the two angular velocity sensors 20 and 21. The threshold is set to be almost constant regardless of the value of the angular velocity.

  The angular velocity sensors 20 and 21 detect an angular velocity as an inertial force, and sensors having the same structure are arranged with the same detection axis. That is, it is mounted redundantly. The angular velocity detection method is not particularly limited. If the angular velocity sensors 20 and 21 having the same configuration have the same sensitivity and offset, the outputs thereof are the same. In the present embodiment, vibration angular velocity sensors are employed as the angular velocity sensors 20 and 21. The first angular velocity sensor 20 corresponds to a first inertial force sensor, and the second angular velocity sensor 21 corresponds to a second inertial force sensor.

  The microcomputer 30 has a function of comparing and diagnosing detection signals input from the angular velocity sensors 20 and 21, and a read-on in which a central processing unit (CPU) (not shown) and various programs executed by the CPU are stored. A memory (ROM), a random access memory (RAM) used as a work area for various calculations executed by the CPU according to each program stored in the ROM, and the like are provided. In the present embodiment, the microcomputer 30 serves as the A / D conversion units 31 and 37, the offset correction unit 32 as the offset correction unit, the correction value calculation unit 33, the correction value setting unit 34, the comparative diagnosis unit 35, and the stop determination unit. The stop determination unit 36 is provided.

  The A / D converter 31 includes an A / D converter that converts analog signals input from the angular velocity sensors 20 and 21 into digital signals. In the present embodiment, the outputs of the two angular velocity sensors 20, 21 are continuously input to one A / D converter by a multiplexer (not shown).

  If the two angular velocity sensors 20 and 21 are also used as an A / D converter, a slight time difference occurs in sampling. It is also conceivable that the difference between the outputs of the two angular velocity sensors 20 and 21 is caused by this time difference, and the determination threshold value at the time of comparative diagnosis is increased accordingly. Therefore, an A / D converter may be provided for each of the angular velocity sensors 20 and 21, and the outputs of the two angular velocity sensors 20 and 21 may be sampled at the same timing.

  The offset correction unit 32 has a function of correcting the detection values of the angular velocity sensors 20 and 21 input via the A / D conversion unit 31 using the offset correction value. This offset correction value is stored in the EEPROM 40. The offset correction unit 32 also has a function of updating the offset correction value based on the detection values of the angular velocity sensors 20 and 21 at this time when receiving the determination signal indicating that the vehicle is stopped from the stop determination unit 36. .

  In this embodiment, the difference between the detected values of the angular velocity sensors 20 and 21 when the vehicle is stopped and the offset correction initial value stored in the EEPROM 40 to cancel the offset at the time of shipment of the angular velocity sensor device 10 is within a predetermined range. For example, the offset correction value is updated to a new value detected this time, and the offset correction value after the update is used in the subsequent offset correction. In addition, the stop determination unit 36 determines the stop of the vehicle based on the detection value of the wheel speed sensor 50 input via the A / D conversion unit 37.

  In addition, the structure of the offset correction | amendment part 32 and the stop determination part 36 is not limited to the said example, A well-known various structure is employable. The stop determination unit 36 may determine whether to stop based on a signal other than the wheel speed sensor 50, for example, a shift position, a hydraulic pressure of a brake device, or the like. Further, for example, as shown in Japanese Patent Application Laid-Open No. 2008-70224 by the present applicant, the offset correction unit 32 does not use other sensor signals such as the wheel speed sensor 50, but based on the signals of the angular velocity sensors 20 and 21. An offset correction may be used.

  As shown in FIG. 3, the angular velocity signals (sensor signals) output from the angular velocity sensor device 10 to the outside (for example, the ECU of the vehicle skid prevention device) are detected by the detection signals of the two angular velocity sensors 20 and 21, as shown in FIG. This is the average value of one or both of the signals that have been A / D converted at 31 and then offset corrected by the offset correction unit 32.

  Only when the detected value of the first angular velocity sensor 20 corrected for offset is a predetermined value (for example, α1) set in advance, the detected values of the angular velocity sensors 20 and 21 corrected for offset are input to the correction value calculation unit 33. Is done. And the correction value calculation part 33 has a function which memorize | stores the frequency | count of a detection for every difference value of the input angular velocity sensors 20 and 21 whenever a detection value is input. Specifically, a histogram representing the relationship between the difference value and the number of detections is created.

  When the histogram is updated, the correction value setting unit 34 extracts a difference value having the highest number of detections (high frequency) from the number of detections of the difference values stored in the correction value calculation unit 33; A function for determining whether or not to rewrite a reference value (hereinafter referred to as the current peak value) stored in the EEPROM 40 to the most frequently extracted difference value (hereinafter referred to as the current peak value) and a condition are as follows. If they match, it has an update function for rewriting the current peak value stored in the EEPROM 40 to the current peak value.

  The comparative diagnosis unit 35 acquires the current peak value stored in the EEPROM 40 and, based on the current peak value, detects the deviation due to the sensitivity difference between the detected values of the angular velocity sensors 20 and 21 input from the offset correction unit 32. It has a function of correcting, and a function of performing a fault diagnosis by comparing a difference value between detection values of the two angular velocity sensors 20 and 21 subjected to offset correction and sensitivity difference correction with a predetermined determination threshold value.

  The EEPROM 40 is a rewritable nonvolatile storage element as is well known, and is a part that stores setting data corresponding to each program. In this embodiment, an offset correction value, a current peak value used for sensitivity difference correction, an offset correction initial value, and the like are stored.

  The offset correction value stored in the EEPROM 40 is input to the offset correction unit 32 when the angular velocity sensors 20 and 21 are activated. Further, the current peak value used for sensitivity difference correction is also input to the comparative diagnosis unit 35 when the angular velocity sensors 20 and 21 are activated.

  Next, among the operations of the angular velocity sensor device 10 according to the present embodiment, a current peak value setting process and a comparative diagnosis process based on the current peak value will be described.

  In the following, the signal of the first angular velocity sensor 20 after the offset correction is Y1a, and the signal of the second angular velocity sensor after the offset correction is Y2a. Further, in this embodiment, the signals Y1a and Y2a of the angular velocity sensors 20 and 21 after offset correction both change linearly with a predetermined inclination with respect to the input angular velocity, whereby the signals Y1a and Y2 of the angular velocity sensors 20 and 21 are changed. The difference of Y2a (Y2a-Y1a) also changes linearly with the signal Y1a of the first angular velocity sensor 20 as the horizontal axis. Furthermore, an example is shown in which the sensitivity (slope) of the signal Y2a of the second angular velocity sensor 21 is greater than the sensitivity (slope) of the signal Y1a of the first angular velocity sensor 20.

  The current peak value setting process shown in the flowchart of FIG. 4 is performed every time the signal Y1a of the first angular velocity sensor 20 corrected for offset becomes a predetermined value (α1) set in advance while the ignition switch is turned on. To be executed. When the ignition switch is turned on, the current peak value for sensitivity difference correction stored in the EEPROM 40 is input to the correction value setting unit 34. Further, the first specified value, the second specified value, and the initial peak value stored in the EEPROM 40 are also input to the correction value setting unit 34.

  When the current peak value setting process is executed, in step 100, the histogram count of the difference value (Y2a-Y1a) of the signals Y1a and Y2a of the angular velocity sensors 20, 21 is updated (number storage means). At this time, the value of the signal Y1a of the first angular velocity sensor 20 is α1.

  FIG. 5 is a diagram showing the difference between the signals Y1a and Y2a of the angular velocity sensors 20 and 21 when the signal Y1a of the first angular velocity sensor 20 is on the horizontal axis. While the vehicle is running, the signals Y1a and Y2a of the angular velocity sensors 20, 21 can take various values as shown in FIG. 5, but in step 100, when the signal Y1a of the first angular velocity sensor 20 is α1, Only the difference value between the signals Y1a and Y2a is acquired.

  Each time the difference value (Y2a−Y1a) is acquired, the histogram is updated as shown in FIG. That is, the acquired difference value is used to update the histogram and constitutes a part of the histogram.

  Further, as the predetermined value α1, a value that can be taken by the signal Y1a of the first angular velocity sensor 20 while the vehicle is turning is set in advance before the angular velocity sensor device 10 is shipped. Preferably, a difference in sensitivity between the two angular velocity sensors 20 and 21 clearly appears (the difference between the signals Y1a and Y2a is large), and a value that can be frequently taken when the vehicle is used is preferably set. Also in the present embodiment, as the predetermined value α1, a difference in sensitivity between the two angular velocity sensors 20 and 21 clearly appears from data at the time of trial production of the angular velocity sensor device 10, and a value that can be frequently taken when the vehicle is used is set. Has been.

  In step 110, as shown in FIG. 6, the difference value βmax that is the maximum number in the histogram updated in step 100 is extracted. The extracted difference value βmax, that is, the current peak value βmax, which is the maximum number of times in the histogram in the current extraction operation, is the difference between the angular velocity sensors 20, 21 when the signal Y1a of the first angular velocity sensor 20 is the predetermined value α1. It is the most frequent value (Y2a-Y1a). Thus, the current peak value βmax is equal to the signal Y2a of the second angular velocity sensor 21 and the first angular velocity sensor when the signal Y1a of the first angular velocity sensor 20 is the predetermined value α1 at the timing at which the predetermined value α1 is acquired this time. This is a shift (difference) due to a difference in sensitivity of the 20 signals Y1a.

  After step 120, it is determined whether or not the current peak value βmax is suitable for updating the current peak value β1. First, in step 120, the current peak value βmax is stored in the EEPROM 40, and a determination is made as to whether or not the current peak value β1 matches the current peak value β1 as the reference value currently used in the comparison diagnosis process (determination means). . The current peak value β1 as the reference value is a value set in the current peak value setting process prior to the current current peak value setting process. That is, the current peak value β1 is also the mode value of the difference value (Y2a−Y1a) when the value of the signal Y1a of the first angular velocity sensor 20 is α1.

  As a result of the determination, if the current peak value βmax matches the current peak value β1, the current peak value setting process is terminated without updating the peak value β1.

  When the current peak value βmax is different from the current peak value β1, in step 130, the difference between the current peak value βmax and the current peak value β1 is compared with a first predetermined value that is set in advance before the angular velocity sensor device 10 is shipped. If it is within the first specified value, it is determined that there is no sudden characteristic change. As the first specified value, a value of about 3% of an initial peak value described later is set in advance.

  On the other hand, when the value exceeds the first specified value, a sudden change due to the addition of the histogram data for one time is abrupt. Then, the current peak value setting process is terminated without updating the current peak value β1. The above corresponds to characteristic sudden change diagnosis means.

  In step 140, the difference between the current peak value βmax and the initial peak value (initial fixed value) set in advance before the shipment of the angular velocity sensor device 10 is a second rule set in advance before the shipment of the angular velocity sensor device 10. If it is within the second specified value, it is determined that there is no change with time. As the second specified value, a value of about 3% of the initial peak value is set in advance.

  On the other hand, if it exceeds the second specified value, it is not a sudden change in characteristics, but it is determined that the change over time has exceeded the allowable range as a result of the change over time (step 145). Then, the current peak value setting process is terminated without updating the current peak value β1. The above corresponds to the time-varying diagnostic means.

  In the present embodiment, an example is shown in which steps 130 and 140 are executed in this order. However, step 140 may be executed first, and then step 130 may be executed.

  Only when it is determined in step 130 and step 140 that there is no failure, in step 150, a rewriting process is performed in which the current peak value βmax is set as a new current peak value β1 (update means). The above is the current peak value setting process.

  The current peak value setting process described above is repeatedly executed every time the signal Y1a of the first angular velocity sensor 20 corrected for offset becomes the predetermined value α1 while the ignition switch is turned on, and the histogram is updated each time. When the ignition switch is turned off, the histogram data created in the correction value calculation unit 33 is deleted. When the ignition is turned on, a new histogram is created in the correction value calculation unit 33.

  Next, the comparative diagnosis process will be described.

  The processing shown in the flowchart shown in FIG. 7 is executed at predetermined processing cycles (for example, every 10 ms) when an ignition switch (not shown) provided in the vehicle is turned on. When the ignition switch is turned on, the offset correction value stored in the EEPROM 40 is input to the offset correction unit 32 and the current peak value β1 for sensitivity difference correction is input to the comparative diagnosis unit 35. Further, the determination threshold value stored in the EEPROM 40 is also input to the comparative diagnosis unit 35. Further, the updated current peak value β1 is input to the comparative diagnosis unit 35 each time the current peak value β1 is updated by the above-described current peak value setting process.

  When the comparative diagnosis process is executed, first, at step 200, an A / D conversion process is executed. That is, the analog signal indicating the detection value from the angular velocity sensors 20 and 21 is converted into a digital signal that can be handled by the microcomputer 30. Hereinafter, the signals of the angular velocity sensors 20 and 21 after A / D conversion are Y1 and Y2, respectively.

  After the A / D conversion process, in step 210, an offset correction process is executed. In this offset correction process, the respective offset correction values are subtracted from the detected values Y1 and Y2 of the angular velocity sensors 20 and 21. Thereby, the detected values of the angular velocity sensors 20 and 21 are offset-corrected. Thereby, the signals of the angular velocity sensors 20 and 21 become Y1a and Y2a described above.

  In step 220, processing for correcting the shift between the two signals Y1a and Y2a due to the sensitivity difference is executed for the signals Y1a and Y2a of the angular velocity sensors 20 and 21 that have been offset corrected.

  Here, since the signals Y1a and Y2a of the angular velocity sensors 20 and 21 are both offset-corrected signals, both are ideally zero when the angular velocity is zero. Accordingly, the difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 when the signal Y1a of the first angular velocity sensor 20 is zero is also zero. The current peak value β1 is the value of the difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 when the signal Y1a of the first angular velocity sensor 20 is the predetermined value α1.

Accordingly, the difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 is expressed as follows, with the signal Y1a of the first angular velocity sensor 20 as the horizontal axis and the difference value (Y2a−Y1a) as the vertical axis, as shown in FIG. 0, 0) and (α1, β1) can be shown (linear approximation) as straight lines passing through two points. That is, the difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 can be expressed by the following equation.
(Expression 1) Y2a−Y1a = (β1 / α1) · Y1a
Thus, the difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 can be expressed as a linear function of the signal Y1a of the first angular velocity sensor 20 having a predetermined inclination (β1 / α1). Therefore, the inclination β1 / α1 corresponds to the difference in sensitivity between the two angular velocity sensors 20, 21. For example, when α1 is 10 ° / s and β1 is 0.2 ° / s, β1 / α1 is 0.02.

  Since α1 for calculating the slope is a fixed value, in the present embodiment, β1 that is the variation value is updated by the current peak value setting process as a reference value for correcting the shift due to the sensitivity difference. I have to. However, β1 / α1 may be used as a reference value for correcting a shift due to a sensitivity difference.

  The signal Y2a of the second angular velocity sensor 21 is obtained by adding the above-described difference value (Y2a−Y1a) to the signal Y1a of the first angular velocity sensor 20. Therefore, the difference due to the difference in sensitivity between the two angular velocity sensors 20 and 21 can be corrected by the difference value (Y2a−Y1a) expressed by Equation 1.

  In this embodiment, the difference (detection value difference) due to the sensitivity difference is corrected by subtracting the above-described difference value (Y2a−Y1a) from the signal Y2a of the second angular velocity sensor 21. In addition, you may correct | amend the shift | offset | difference by a sensitivity difference by adding above-mentioned difference value (Y2a-Y1a) to the signal Y1a of the 1st angular velocity sensor 20. FIG.

In step 230, the difference between the detection values of the two angular velocity sensors 20 and 21 that have been subjected to the sensitivity difference correction is compared with a determination threshold value that is set in advance before shipment of the angular velocity sensor device 10. This judgment formula can be expressed by the following formula.
(Equation 2) | {Y2a− (β1 / α1) · Y1a} −Y1a |> Determination Threshold Value When the above Equation 2 is not satisfied, that is, the difference between the detected values of the two angular velocity sensors 20 and 21 corrected for sensitivity difference If it is within the determination threshold, it is determined that there is no failure (step 240), and if Equation 2 is satisfied, that is, if the determination threshold is exceeded, it is determined as a one-side failure (step 250). These determination results are output to the outside as a diagnosis signal.

  Note that the above-described steps 220 and 230 correspond to comparative diagnosis means. Equation 2 above is stored in the ROM, and the current peak value β1 input from the EEPROM 40 and the signals Y1a and Y2a of the angular velocity sensors 20 and 21 that have been offset corrected are applied to Equation 2 to correct the sensitivity difference (step 220). At the same time, a comparative diagnosis (step 230) is executed.

  Thus, in the angular velocity sensor device 10 according to the present embodiment, a histogram is created based on the difference value (Y2a−Y1a) between the two angular velocity sensors 20 and 21 mounted redundantly, and the maximum number of times is obtained in the histogram. The difference value βmax (β1) is used for correcting the deviation of the detection value due to the sensitivity difference. Therefore, when there is no failure such as a short circuit, the difference value between the angular velocity sensors 20 and 21 after the sensitivity difference correction is calculated as shown in FIG. 9 (the value of the signal Y1a of the first angular velocity sensor 20). Regardless of this, it can be set to a substantially constant value.

  As a result, the determination threshold can be set to a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1a of the first angular velocity sensor 20), as shown in FIG. That is, the determination threshold can be reduced even when the angular velocity value is large as compared with the conventional case. Thereby, the control precision of skid prevention can be improved.

(Modification)
As shown in FIG. 10, even when the difference (Y2a−Y1a) between the signals Y1a and Y2a of the angular velocity sensors 20 and 21 with respect to the signal Y1a of the first angular velocity sensor 20 is non-linear, the detected value due to the sensitivity difference The concept of correcting the deviation is the same as that in the first embodiment.

  For example, similarly to the above, the difference due to the sensitivity difference may be corrected by subtracting the difference value (Y2a−Y1a) from the signal Y2a of the second angular velocity sensor 21.

  However, since the difference value (Y2a−Y1a) is non-linear, a plurality of predetermined values of the signal Y1a of the first angular velocity sensor 20 are set, and the non-linearity is estimated by approximation.

  For example, in FIG. 10, four points α2 to α5 are set in advance as predetermined values. Α2 and α3 (<α2) are positive values, α4 and α5 (<α4) are negative values, α2 and α5 have the same absolute value, and α3 and α4 have the same absolute value. . Then, as with the above-described predetermined value α1, for each predetermined value α2, α3, α4, α5, the corresponding difference value (Y2a-Y1a) histogram is updated each time a corresponding value is input. Then, the current peak values β2 to β5 are set.

  Further, in the comparative diagnosis process, the signal Y1a of the first angular velocity sensor 20 is the horizontal axis and the difference value (Y2a−Y1a) is the vertical axis, and (α2, β2), (α3, β3), (α4, β4), ( The difference value (Y2a−Y1a) between the angular velocity sensors 20 and 21 is approximated as a function of the signal Y1a of the first angular velocity sensor 20 from the four points data of α5 and β5). This approximate calculation is executed in step 220 described above. Further, programs for approximation calculation processing (polynomial approximation processing, exponential approximation processing, etc.) are stored in the ROM.

For example, when the difference value (Y2a−Y1a) is a quadratic function, the following equation is obtained.
(Expression 3) Y2a−Y1a = c · Y1a ² + d · Y1a
Here, c and d are fixed values.

Then, the determination formula of the difference between the detection values of the two angular velocity sensors 20 and 21 corrected for sensitivity difference is as follows.
(Equation 4) | {Y2a− (c · Y1a ² + d · Y1a)} − Y1a |> determination threshold As described above, even when the difference value (Y2a−Y1a) is non-linear, the angular velocity sensor after the sensitivity difference correction is performed. The difference value between 20 and 21 can be set to a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1a of the first angular velocity sensor 20). The determination threshold can also be set to a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1a of the first angular velocity sensor 20).

  In order to approximate the non-linear difference value (Y2a−Y1a) as a function of the signal Y1a of the first angular velocity sensor 20, data of three or more points is necessary. However, as described above, when the signal Y1a of the angular velocity sensors 20, 21 is zero, the difference value (Y2a-Y1a) is also zero. Therefore, when (0, 0) is used, at least two points of data are sufficient. For example, the above (α2, β2) and (α4, β4) and (0, 0) may be approximated together. Note that as the predetermined value for updating the histogram is increased, the approximation accuracy of the nonlinearity increases, and the sensitivity difference can be corrected with high accuracy for the nonlinearity.

(Second Embodiment)
Only the parts different from the first embodiment will be described below. In the present embodiment, as shown in FIG. 11, the microcomputer 30 includes A / D conversion units 31 and 37, an offset correction unit 32, an offset difference / sensitivity difference calculation unit 38, a correction value setting unit 34, a comparative diagnosis unit 35, And a stop determination unit 36. That is, an offset difference / sensitivity difference calculation unit 38 is provided instead of the correction value calculation unit 33.

  The offset difference / sensitivity difference calculation unit 38 receives the detection values (raw values not subjected to offset processing) of the angular velocity sensors 20 and 21 subjected to A / D conversion. The offset difference / sensitivity difference calculation unit 38 stores the difference value of the input detection value, and each time the difference value is stored, the stored difference value is compared with the detection value of the angular velocity sensor 20. It has a function of calculating an approximate line by approximation. Furthermore, it has a function for obtaining a calculated value for sensitivity difference correction and a calculated value for offset difference correction as a corrected calculated value (hereinafter, referred to as a current calculated value) from the calculated approximate line.

  In the present embodiment, when the correction value setting unit 34 updates the current calculated value, it is determined whether or not the correction value stored in the EEPROM 40 (hereinafter referred to as the current correction value) is rewritten to the current calculated value. The determination function and the update function for rewriting the current correction value stored in the EEPROM 40 to the current calculated value when the condition is met.

  In addition, the comparative diagnosis unit 35 acquires the current correction value stored in the EEPROM 40, and based on the current correction value, the difference between the offset difference and the sensitivity difference between the detected values of the angular velocity sensors 20, 21 subjected to A / D conversion. And a function of performing a fault diagnosis by comparing a difference value between detection values of the two angular velocity sensors 20 and 21 subjected to offset difference correction and sensitivity difference correction with a predetermined determination threshold value.

  The EEPROM 40 stores a current correction value (current correction value for offset difference correction and current correction value for correction), an offset correction value used in the offset correction unit 32, an offset correction initial value, and the like. The current correction value stored in the EEPROM 40 is input to the comparative diagnosis unit 35 when the angular velocity sensors 20 and 21 are activated.

  Next, among the operations of the angular velocity sensor device 10 according to the present embodiment, a current correction value setting process and a comparative diagnosis process based on the current correction value will be described.

  In the following, the signal of the first angular velocity sensor 20 after A / D conversion is Y1, and the signal of the second angular velocity sensor 21 after A / D conversion is Y2. In the present embodiment, the signals Y1 and Y2 of the angular velocity sensors 20 and 21 after A / D conversion both change linearly with a predetermined inclination with respect to the input angular velocity, whereby the signals of the angular velocity sensors 20 and 21 are obtained. The difference between Y1 and Y2 (Y2−Y1) also changes linearly with the signal Y1 of the first angular velocity sensor 20 as the horizontal axis. Furthermore, an example is shown in which the sensitivity (slope) of the signal Y2 of the second angular velocity sensor 21 is greater than the sensitivity (slope) of the signal Y1 of the first angular velocity sensor 20.

  The current correction value setting process shown in the flowchart shown in FIG. 12 is executed every predetermined processing cycle (for example, every processing cycle longer than the comparative diagnosis processing) when the ignition switch is turned on. When the ignition switch is turned on, the current correction values (the current correction value for offset difference correction and the current correction value for sensitivity difference correction) stored in the EEPROM 40 are input to the correction value setting unit 34. Further, the first specified value, the second specified value, and the initial correction value stored in the EEPROM 40 are also input to the correction value setting unit 34.

  When the current correction value setting process is executed, in step 300, the difference value (Y2−Y1) between the signals Y1 and Y2 of the angular velocity sensors 20 and 21 is stored. Each time it is stored, a plurality of difference values (Y2−Y1) stored so far including the difference value stored this time are approximated as a function of the signal Y1 of the first angular velocity sensor 20, and thus an approximate line Is calculated.

In the present embodiment, as shown in FIG. 13, a plurality of stored difference values (Y2−Y1) are linearly approximated. By this linear approximation, if the horizontal axis is the signal Y1 of the first angular velocity sensor 20 and the vertical axis is the difference (Y2-Y1) between the signals Y1 and Y2 of the angular velocity sensors 20, 21, the difference (Y2-Y1) is It will be shown as a linear function of the signal Y1 of the angular velocity sensor 20.
(Expression 5) Y2−Y1 = e1 · Y1 + f1
In Equation 5, the inclination e1 corresponds to the difference in sensitivity between the signals Y1 and Y2 of the angular velocity sensors 20 and 21, and the intercept f1 corresponds to the difference in offset between the signals Y1 and Y2 of the angular velocity sensors 20, 21.

  In step 300, the slope e1 (currently calculated value of sensitivity difference) and the intercept f1 (currently calculated value of offset difference) are calculated (correction value calculating means) as the current calculated value. The linear approximation and the calculation of the current calculated value are executed by a program stored in the ROM.

  As described above, in the present embodiment, the calculated values of the sensitivity difference and the offset difference are obtained by approximate calculation without using a histogram.

  Step 310 and subsequent steps are basically the same as step 120 and subsequent steps in the current peak value setting process shown in the first embodiment. In other words, the currently calculated value calculated this time (the current calculated value of the sensitivity difference and the current calculated value of the offset difference) are stored in the EEPROM 40 (the current corrected value of the sensitivity difference and the current corrected value of the offset difference). It is determined whether each is suitable for updating. The current correction value for the sensitivity difference corresponds to the sensitivity difference correction value, and the current correction value for the offset difference corresponds to the offset difference correction value.

  First, in step 310, the currently calculated values (e1, f1) are stored in the EEPROM 40 and are used in the current comparative diagnosis process (current correction value e0 for sensitivity difference and current correction value f0 for offset difference). Determination of whether each corresponds is performed (determination means). The current correction values (the current correction value e0 for the sensitivity difference and the current correction value f0 for the offset difference) are values set in the current correction value setting process before the current calculated values (e1, f1). That is, the current correction value (e0, f0) is also data of the slope and intercept of a linear function obtained by linearly approximating a plurality of difference values (Y2−Y1).

  If the current calculated value (e1, f1) matches the corresponding current correction value (e0, f0) as a result of the determination, the current correction value (e0, f0) is not updated and the current correction value is updated. The value setting process ends. If only one of the current calculated values (e1, f1) matches the corresponding current correction value (e0, f0), the current correction value on the matching side (for example, sensitivity difference) is updated. Instead, the current correction value setting process ends. On the other hand, for the non-matching side (for example, offset difference), the steps after step 320 are executed.

  If at least one of the current calculated values (e1, f1) is different from the corresponding current correction value (e0, f0), in step 320, the difference between the different current calculated value and the current correction value is the difference between the angular velocity sensor device 10 and the current calculated value. It is compared with a first specified value set in advance before shipment, and if it is within the first specified value, it is determined that there is no sudden characteristic change.

  On the other hand, when the value exceeds the first specified value, a change due to the addition of the difference value data for one time is abrupt, so that it is determined that the characteristic sudden change failure (step 325). Then, the current correction value setting process is terminated without updating the current correction value. The above corresponds to characteristic sudden change diagnosis means.

  In step 330, the difference between the current calculated value that does not match the current correction value and the corresponding initial correction value set in advance before the shipment of the angular velocity sensor device 10 is set to a first value set in advance before the shipment of the angular velocity sensor device 10. 2 is compared with a specified value, and if it is within the second specified value, it is determined that there is no change with time.

  On the other hand, if it exceeds the second specified value, it is not a sudden characteristic change, but it is determined that a time-dependent failure has occurred because the time-dependent changes have accumulated and exceeded the allowable range (step 335). Then, the current correction value setting process is terminated without updating the current correction value. The above corresponds to the time-change diagnosis means.

  In the present embodiment, an example is shown in which steps 320 and 330 are executed in this order. However, step 330 may be executed first, and then step 320 may be executed.

  Only when it is determined that there is no failure in step 320 and step 330, in step 340, a rewriting process is performed in which the currently calculated values (for example, e1 and f1) are set as new current correction values (for example, e0 and f0). (Update means). The above is the current correction value setting process.

  The current correction value setting process described above is repeatedly executed while the ignition switch is turned on, and the approximate line is updated each time. When the ignition switch is turned off, the difference value data stored in the offset difference / sensitivity difference calculation unit 38 is deleted. When the ignition is turned on, a new difference value is stored in the offset difference / sensitivity difference calculation unit 38, and an approximate line is calculated based on the new difference value.

  Next, the comparative diagnosis process will be described.

  The process shown in the flowchart of FIG. 14 is executed at predetermined processing cycles (for example, every 10 ms) when an ignition switch (not shown) provided in the vehicle is turned on. When the ignition switch is turned on, the current correction values (the current sensitivity correction value e0 for the sensitivity difference and the current correction value f0 for the offset difference) stored in the EEPROM 40 are input to the comparison diagnosis unit 35. Further, the determination threshold value stored in the EEPROM 40 is also input to the comparative diagnosis unit 35. In addition, the current correction value is input to the comparative diagnosis unit 35 each time the value is updated by the above-described current correction value setting process.

  When the comparative diagnosis process is executed, first, at step 400, an A / D conversion process is executed. Then, after the A / D conversion process, in step 410, a process for correcting the deviation due to the offset difference and the sensitivity difference is performed on the signals Y1 and Y2 of the angular velocity sensors 20 and 21.

  Here, the signal Y2 of the second angular velocity sensor 21 is obtained by adding the difference value (Y2-Y1) shown in the above equation 5 to the signal Y1 of the first angular velocity sensor 20. Therefore, the deviation due to the difference in sensitivity between the two angular velocity sensors 20 and 21 and the deviation due to the offset difference can be corrected by the difference value (Y2−Y1) expressed by Equation 5.

  In the present embodiment, the difference due to the sensitivity difference and the deviation due to the offset difference are corrected by subtracting the above-described difference value (Y2-Y1) from the signal Y2 of the second angular velocity sensor 21. In addition, you may correct | amend the shift | offset | difference by a sensitivity difference, and the shift | offset | difference by an offset difference by adding above-mentioned difference value (Y2-Y1) to the signal Y1 of the 1st angular velocity sensor 20. FIG.

In step 420, the difference between the detection values of the two angular velocity sensors 20, 21 corrected for the deviation due to the sensitivity difference and the deviation due to the offset difference is compared with a determination threshold set in advance before shipment of the angular velocity sensor device 10. This judgment formula can be expressed by the following formula.
(Equation 6) | {Y2− (e1 · Y1 + f1)} − Y1 |> Determination Threshold Value When the above Equation 6 is not satisfied, that is, the detection values of the two angular velocity sensors 20 and 21 in which the sensitivity difference and the offset difference are corrected. If the difference is within the determination threshold, it is determined that there is no failure (step 430), and if Equation 6 is satisfied, that is, if the determination threshold is exceeded, it is determined as a one-side failure (step 440). These determination results are output to the outside as a diagnosis signal.

  The above-described steps 410 and 420 correspond to a comparative diagnosis unit. Equation 6 is stored in the ROM, and the current correction values (e0, f0) input from the EEPROM 40 and the signals Y1, Y2 of the angular velocity sensors 20, 21 are applied to Equation 6, and sensitivity difference and offset difference correction ( A comparative diagnosis (step 420) is executed together with step 410).

  As described above, in the angular velocity sensor device 10 according to the present embodiment, instead of creating a histogram and extracting a peak value as shown in the first embodiment, approximation is performed from a plurality of stored difference values (Y2−Y1). The approximate line is calculated, and the coefficients (e1, f1) of the approximate line are used for correcting the deviation of the detected value due to the sensitivity difference and the offset difference. As a result, as in the first embodiment, the difference value between the two angular velocity sensors 20 and 21 in which the sensitivity difference and the offset difference are corrected is substantially the same regardless of the magnitude of the angular velocity (the value of the signal Y1 of the first angular velocity sensor 20). It can be a constant value.

  The determination threshold can also be set to a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1 of the first angular velocity sensor 20). That is, the determination threshold can be reduced even when the angular velocity value is large as compared with the conventional case. Thereby, the control precision of skid prevention can be improved.

(Modification)
As shown in FIG. 15, the change in the difference (Y2−Y1) between the signals Y1 and Y2 of the angular velocity sensors 20 and 21 with respect to the signal Y1 of the first angular velocity sensor 20 is also non-linear, depending on the sensitivity difference and the offset difference. The concept of correcting the deviation of the detected value is the same as in the second embodiment.

  For example, similarly to the above, the difference due to the sensitivity difference may be corrected by subtracting the difference value (Y2−Y1) from the signal Y2 of the second angular velocity sensor 21.

For example, in FIG. 15, a plurality of difference values (Y2−Y1) are approximated by a polynomial, whereby the horizontal axis indicates the signal Y1 of the first angular velocity sensor 20 and the vertical axis indicates the difference between the signals Y1 and Y2 of the angular velocity sensors 20 and 21 ( As Y2-Y1), the difference value (Y2-Y1) is shown as an n-order function (n ≧ 2) of the signal Y1 of the first angular velocity sensor 20. For example, when the difference value (Y2−Y1) is a quadratic function, the following equation is obtained.
(Expression 7) Y2−Y1 = g1 · Y1 ² + h1 · Y1 + j1
In Equation 7, g1 and h1 are gradient data corresponding to the difference in sensitivity between the signals Y1 and Y2 of the angular velocity sensors 20 and 21, and j1 is an intercept corresponding to the difference in offset between the signals Y1 and Y2 of the angular velocity sensors 20 and 21. . These g1, h1, and j1 are the calculated values this time. The approximation and the calculation of the current calculated value are executed by a program stored in the ROM.

Then, the determination formula of the difference between the detection values of the two angular velocity sensors 20, 21 in which the sensitivity difference and the offset difference are corrected by the current correction values g0, h0, j0 is as follows.
(Equation 8) | {Y2− (g0 · Y1 ² + h0 · Y1 + j0)} − Y1 |> Determination Threshold As described above, even when the difference value (Y2−Y1) is non-linear, the sensitivity difference and the offset difference are corrected. The difference value between the angular velocity sensors 20 and 21 can be a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1 of the first angular velocity sensor 20). The determination threshold can also be set to a substantially constant value regardless of the magnitude of the angular velocity (the value of the signal Y1 of the first angular velocity sensor 20).

  As an approximation method, a preferable one may be set in advance as appropriate from the change of the difference value (Y2−Y1) with respect to the signal Y1 of the angular velocity sensor 20. For example, in addition to polynomial approximation, exponential approximation or the like can be used.

  The preferred embodiments of the present invention have been described above. However, the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention.

  In this embodiment, the example of the angular velocity sensor was shown as an inertial force sensor. However, the present invention can also be applied to an acceleration sensor that detects acceleration as an inertial force. Furthermore, in an apparatus including a pair of angular velocity sensors and a pair of acceleration sensors, the present invention can be applied to each of the angular velocity sensor and the acceleration sensor.

10 ... Angular velocity sensor device (Inertial force sensor device)
20 ... 1st angular velocity sensor (1st inertial force sensor)
21... Second angular velocity sensor (second inertial force sensor)
31 ... A / D conversion unit 32 ... offset correction unit 33 ... correction value calculation unit 34 ... correction value setting unit 35 ... comparative diagnosis unit 38 ... offset difference / sensitivity difference calculation unit 40... EEPROM (correction value storage means)

Claims (5)

A first inertial force sensor and a second inertial force sensor of the same configuration that are provided in a vehicle and output a detection signal corresponding to the inertial force;
The inertial force value corresponding to the detection signals of the two inertial force sensors and the inertial force value after offset correction is input, and the inertial force value of the first inertial force sensor that has been offset corrected is set in advance. A number-of-times storage means for storing the number of times of detection for each difference value of the inertial force values of the two inertial force sensors that are input each time a predetermined value is obtained when the inertial force acts on the vehicle;
Non-volatile correction value storage means that is rewritable and stores the reference value;
A determination unit that extracts a difference value having the largest number of detections from the number storage unit and determines whether or not the extracted value matches the reference value;
When the determination means determines that the extracted value does not match the reference value, the extracted value is compared with the reference value, and a difference between the extracted value and the reference value is preset. When the first specified value is exceeded, the characteristic sudden change diagnosis means for determining that there is a failure;
When the determination means determines that the extracted value does not match the reference value, the extracted value is compared with a preset initial fixed value, and the difference between the extracted value and the initial fixed value is compared. When the time exceeds a second predetermined value set in advance, a time-change diagnosis means for determining that a failure has occurred;
An update means for rewriting the reference value stored in the correction value storage means to the extracted value when it is determined that there is no failure in the characteristic sudden change diagnosis means and the temporal change diagnosis means;
Based on the reference value stored in the correction value storage means, the sensitivity difference between the inertial force values of the two inertial force sensors after offset correction is corrected, and the two inertial force sensors subjected to offset correction and sensitivity difference correction are corrected. The inertial force sensor device comprising: a comparison diagnosis unit that determines that one of the failure is a failure when a difference value between the inertial force values exceeds a predetermined determination threshold value. The two inertial force sensors are angular velocity sensors,
An offset correction unit that receives an angular velocity value corresponding to detection signals of the two angular velocity sensors and corrects an offset of the angular velocity value;
Stop determination means for determining that the vehicle is stopped, and
The offset correction means offset-corrects the angular velocity values of the two angular velocity sensors based on the angular velocity values of the two angular velocity sensors input when the vehicle determination unit determines that the vehicle is stopped. The inertial force sensor device according to claim 1. The number-of-times storage means stores the number of detections for each difference value according to each predetermined value, with a plurality of the predetermined values.
The inertial force sensor device according to claim 1 or 2, wherein the reference value is stored for each of the predetermined values in the correction value storage means. A first inertial force sensor and a second inertial force sensor of the same configuration that are provided in a vehicle and output a detection signal corresponding to the inertial force;
A difference value between inertial force values corresponding to the detection signals input from the two inertial force sensors is stored, and each time the difference value is stored, a plurality of stored difference values are stored in the first difference value. Approximate an approximate line in relation to the inertial force value of the inertial force sensor, and based on the approximate line, calculate a correction value for calculating a sensitivity difference correction calculation value and an offset difference correction calculation value as a correction calculation value Means,
Non-volatile correction value storage means that is rewritable and stores sensitivity difference correction values and offset difference correction values as correction values;
Determination means for determining whether or not the correction calculation value and the corresponding correction value match each other;
When the determination means determines that the correction calculation value does not match the corresponding correction value, and the difference between the correction calculation value and the correction value exceeds a preset first predetermined value , Characteristic sudden change diagnostic means for determining that there is a failure,
When the determination means determines that the correction calculation value does not match the correction value, the correction calculation value is compared with a preset initial fixed value of correction, and the correction calculation value and the initial value are compared. When the difference from the fixed value exceeds the corresponding second predetermined value set in advance, the time-change diagnosis means for determining that there is a failure;
Updating means for rewriting the correction value stored in the correction value storage means to the correction calculated value determined as having no failure when it is determined that there is no failure in the characteristic sudden change diagnosis means and the temporal change diagnosis means; ,
Based on the correction value stored in the correction value storage means, the offset difference and sensitivity difference between the inertial force values of the two inertial force sensors are corrected, and the two inertial force sensors subjected to offset difference correction and sensitivity difference correction are corrected. The inertial force sensor device comprising: a comparison diagnosis unit that determines that one of the failure is a failure when a difference value between the inertial force values exceeds a predetermined determination threshold value.   The inertial force sensor device according to claim 4, wherein the two inertial force sensors are angular velocity sensors.

Images