JP2019066318A - Electronic equipment of vehicle - Google Patents

Abstract

To complete processing to be executed according to a plurality of pieces of acquired data at an early stage.SOLUTION: Electronic equipment of a vehicle comprises: an acquisition unit that acquires data; a determination unit that, prior to a determination of whether a variation degree of a first prescribed number of data is less than a first prescribed value, determines whether a probability in which the variation degree of the first prescribed number of data is equal to or more than the first prescribed value is equal to or more than a second prescribed value on the basis of the second prescribed number of data at a point of time of acquisition of the second prescribed number of data smaller than the first prescribed number thereof by the acquisition unit, and when determining that the probability is equal to or more than the second prescribed value, conducts reset processing for selecting new data without using all or a part of the second prescribed number of data; and a data processing unit that, when it is determined by the determination unit that the variation degree of the first prescribed number of data is less than the first prescribed value, implements prescribed processing, using the first prescribed number of data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an electronic device of a vehicle.

  Conventionally, various sensors such as an acceleration sensor and a yaw rate sensor are mounted on a vehicle. In addition, an electronic device that performs processing based on data of various sensors is mounted. For example, some electronic devices perform a zero point correction process that corrects a reference zero point based on a predetermined number of sample data obtained from a sensor. When correcting the zero point, if the degree of variation of the predetermined number of sample data acquired from the sensor is large, the reliability of the processing result is reduced, so it is necessary to reacquire the sample data from the sensor.

Unexamined-Japanese-Patent No. 6-3368

  If the degree of variation is determined after acquiring a predetermined number of sample data and the sample data is acquired again as necessary, the completion of the process will be delayed.

  Therefore, one of the problems of the present invention is, for example, to complete early the process performed based on the plurality of acquired data.

  The present invention relates to, for example, an acquisition unit for acquiring data, and the first acquisition unit for determining whether or not the degree of variation of the first predetermined number of data is less than a first predetermined value. When a second predetermined number of the data smaller than the predetermined number is acquired, the degree of variation of the first predetermined number of the data is determined based on the second predetermined number of data. It is determined whether the probability of becoming equal to or greater than the value is equal to or greater than a second predetermined value, and when it is determined that the probability is equal to or greater than the second predetermined value, all or one of the second predetermined number of data A determination unit that performs a reset process to select new data without using a unit, and the determination unit determines that the variation degree of the first predetermined number of data is less than the first predetermined value, Perform predetermined processing using the first predetermined number of data A data processing unit that is an electronic apparatus of a vehicle including a.

  Further, in the zero point correction device, for example, the determination unit may determine that the first predetermined number is N, the second predetermined number is S, and the S pieces of data have a variation of 0 (zero The reset process is performed on the basis of the result of the variation degree calculated based on the N-S pieces of data assumed to be.

In the zero point correction apparatus, for example, the first predetermined number is N, the data is x _i (1 ≦ i ≦ N), the first predetermined value is E, and the second predetermined number is S. (2 ≦ S <N), the second predetermined value is set to 100%, and the determination unit performs the reset process when the following formula (11) is not satisfied.

  Further, in the zero point correction device, for example, the determination unit determines that the second predetermined number is S, and the degree of variation of the first predetermined number of the data is determined based on the S pieces of data. A determination value for determining whether the probability of becoming equal to or more than the predetermined value of 1 becomes equal to or more than the second predetermined value is calculated based on the chi-square distribution of the degree of freedom S-1.

In the zero point correction apparatus, for example, the first predetermined number is N, the data is x _i (1 ≦ i ≦ N), the first predetermined value is E, and the second predetermined number is S. (2 ≦ S <N), when the third predetermined value corresponding to the second predetermined value is E4, the determination unit performs the reset process when the following equation (12) is not satisfied: Do.

  Further, in the zero point correction apparatus, for example, after the second predetermined number of data are acquired, the determination unit calculates an average value of the data acquired by then, and the average value of the data is calculated from the average value. The data having the deviation degree equal to or more than the fourth predetermined value is discarded.

  Further, in the zero point correction device, for example, when the number of discards is equal to or more than a predetermined number, the determination unit performs a reset process of selecting new data without using all of the second predetermined number of data. .

  Further, in the zero point correction device, for example, the acquisition unit acquires data detected by a sensor, and the data processing unit uses the first predetermined number of data to obtain the reference value of the sensor. Implement a zero point correction to correct a certain zero point.

FIG. 1 is a view showing a schematic configuration of a vehicle equipped with an ECU (electronic device of the vehicle) of the embodiment. FIG. 2 is a flowchart showing the procedure of a first zero point correction process by the ECU of the embodiment. FIG. 3 is a graph showing the relationship between the calculated value of data and the threshold value E in the embodiment. FIG. 4 is a graph showing a flow of zero point correction processing when the method of RMS2 and the method of RMS3 are used in the embodiment. FIG. 5 is a graph showing the flow of the zero point correction process when the yaw rate sensor value is acquired and the method of RMS2 is used and the method of RMS3 is used in the embodiment. FIG. 6 is a graph showing the relationship between the index value Z and the threshold value E4 in the embodiment. FIG. 7 is an explanatory view of a normal distribution and the like. FIG. 8 is a flowchart showing the procedure of a second zero point correction process by the ECU of the embodiment.

  In the following, exemplary embodiments of the present invention are disclosed. The configurations of the embodiments shown below, and the operations and results (effects) provided by the configurations are examples. The present invention can also be realized with configurations other than the configurations disclosed in the following embodiments. Further, according to the present invention, it is possible to obtain at least one of various effects (including derivative effects) obtained by the following configuration.

  First, with reference to FIG. 1, a schematic configuration of a vehicle equipped with an ECU (Electronic Control Unit) 50 (electronic device of the vehicle) of the embodiment will be described. FIG. 1 is a view showing a schematic configuration of a vehicle equipped with the ECU 50 of the embodiment.

  As shown in FIG. 1, in the vehicle, a driving force generator 11 for causing the vehicle to travel, a steering device 12 for turning the front wheels FR, FL as steered wheels, and the wheels FL, FR, RL, A braking device 13 is provided to apply a braking force to RR (hereinafter sometimes referred to simply as "wheel" with reference numerals omitted).

  The driving force generation device 11 is provided with an engine 21 that generates a driving force based on an operation amount of the accelerator pedal 20 by the driver, that is, an accelerator opening degree, and an automatic transmission 22 connected to an output shaft of the engine 21. It is done. Further, the driving force generation device 11 is provided with an accelerator opening degree sensor SE1 for detecting an accelerator opening. The driving force output from the engine 21 is transmitted from the automatic transmission 22 to the differential gear 23, and is distributed from the differential gear 23 to the front wheels FR and FL which are driving wheels.

  The vehicle is provided with a shift device 25 operated by the driver. When the range of the shift device 25 is the forward range, the automatic transmission 22 outputs a driving force in the direction of advancing the vehicle. On the other hand, when the range of the shift device 25 is the reverse range, the automatic transmission 22 outputs a driving force in a direction to reverse the vehicle. Shift information, such as whether it is the forward range or the reverse range, is transmitted to the ECU 50 that controls the braking device 13.

  The steering device 12 is provided with a steering shaft 31 to which a steering wheel 30 steered by a driver is fixed, and a steering actuator 32 connected to the steering shaft 31. Further, the steering device 12 is provided with a link mechanism portion 33 including a tie rod movable in the lateral direction of the vehicle by the steering actuator 32, and a link for steering the front wheels FL and FR by the movement of the tie rods. . The steering device 12 is also provided with a steering angle sensor SE2 that outputs a detection signal corresponding to the steering angle of the steering 30 to the ECU 50. A steering angle sensor SE2 outputs, for example, a detection signal that makes the steering angle a positive value when turning the vehicle to the right, and a steering angle of turning the vehicle to the left. It outputs a detection signal that has a negative value.

  The braking device 13 includes a hydraulic pressure generating device 41 that generates a brake hydraulic pressure according to the operating force of the brake pedal 40 by the driver, and a braking device 42a individually provided for each of the wheels FR, FL, RR, and RL A brake actuator 43 connected to 42b, 42c and 42d is provided. The braking device 13 is also provided with a brake switch SW1 that outputs a detection signal to the ECU 50 according to the operating condition (on / off) of the brake pedal 40 by the driver.

  The brake actuator 43 is configured to be able to apply a braking force to each of the wheels FR, FL, RR, and RL even when the driver does not operate the brake pedal 40. For example, the brake actuator 43 generates a differential pressure between the brake fluid pressure on the fluid pressure generating device 41 side and the brake fluid pressure in the wheel cylinders provided in the brake devices 42a, 42b, 42c, 42d. A differential pressure control valve and an electric pump for supplying brake fluid into the wheel cylinder are provided. In addition, the brake actuator 43 is provided with various valves for individually adjusting the brake fluid pressure in each wheel cylinder. That is, the brake actuator 43 of this embodiment can individually adjust the braking force on each of the wheels FR, FL, RR, and RL.

  In addition to the accelerator opening sensor SE1, the steering angle sensor SE2, and the brake switch SW1, the ECU 50 also includes wheel speed sensors SE3, SE4, SE5, SE6 for detecting the wheel speeds of the respective wheels FR, FL, RR, RL. Are electrically connected. Further, the ECU 50 includes a longitudinal acceleration sensor SE7 for detecting an acceleration in the longitudinal direction of the vehicle (hereinafter referred to as "longitudinal acceleration") and an acceleration (hereinafter referred to as "lateral direction of the vehicle"). A lateral acceleration sensor SE8 for detecting a lateral acceleration) and a yaw rate sensor SE9 for detecting a yaw rate of the vehicle are electrically connected.

  When the road surface on which the vehicle travels is a horizontal road, the longitudinal acceleration sensor SE7 outputs a detection signal such that the longitudinal acceleration has a positive value when the vehicle accelerates, and the vehicle decelerates. A detection signal is output such that the longitudinal acceleration has a negative value. Further, when the vehicle stops on a downhill road, the center of gravity of the vehicle moves to the front side, so the longitudinal acceleration calculated based on the detection signal from the longitudinal acceleration sensor SE7 has a negative value.

  Also, from the lateral acceleration sensor SE8 and the yaw rate sensor SE9, when the vehicle turns in the right direction, the lateral acceleration and the yaw rate of the vehicle become positive values, while when the vehicle turns in the left direction Detection signals are output such that the lateral acceleration and the yaw rate of the vehicle become negative values.

  The ECU 50 may be incorporated into an ECU of any system mounted on a vehicle, or may be an independent ECU. The ECU 50 includes, for example, a central processing unit (CPU) (not shown), a controller, a random access memory (RAM), a read only memory (ROM), a flash memory, and the like. The ECU 50 can execute processing in accordance with the installed and loaded program to realize each function.

  The ECU 50 includes a processing unit 510 and a storage unit 520. The processing unit 510 includes an acquisition unit 511, a determination unit 512, and a zero point correction unit 513 which is an example of a data processing unit that performs a predetermined process. That is, the ECU 50 can function as the acquisition unit 511, the determination unit 512, the zero point correction unit 513, and the like by executing the process according to the program. The storage unit 520 also stores data used in the calculation process of each unit, data of the result of the calculation process, and the like. Note that at least a part of the functions of the above-described units may be realized by hardware.

  The acquisition unit 511 acquires data (detection data) from the sensors SE1 to SE9 mounted on the vehicle, the brake switch SW1, and the like.

  Among the sensors SE1 to SE9, for example, a yaw rate sensor SE9 is an object to be corrected of the zero point. The zero point is the reference value of the sensor. The zero point is, for example, in the case of the yaw rate of the vehicle, a value that indicates that the yaw rate of the vehicle is zero, that is, the vehicle is not rotating at all. The yaw rate sensor SE9 may have a difference between the currently set zero value and the actual zero value due to temperature drift or the like, which is a change in output voltage or the like due to a change in ambient temperature. Therefore, it is necessary to correct the zero point. In the zero point correction of the yaw rate sensor SE9, for example, the data (sample data) of the yaw rate sensor SE9 when the vehicle is traveling straight is used to perform the zero point correction (details will be described later). In the following, an object for correcting the zero point is simply referred to as a "sensor". Also, sample data may be simply referred to as "data".

  The determination unit 512 determines whether the degree of variation of the first predetermined number of data is less than the first predetermined value. Before the determination unit 512 obtains the second predetermined number of data smaller than the first predetermined number by the acquisition unit 511, the determination unit 512 determines the first predetermined number based on the second predetermined number of data. If it is determined whether the probability that the degree of variation in the number data is greater than or equal to the first predetermined value is greater than or equal to the second predetermined value, and it is determined that the probability is greater than or equal to the second predetermined value, the second And reset processing for selecting new data without using all or part of the predetermined number of data (details will be described later).

  If the determination unit 512 determines that the variation degree of the first predetermined number of data is less than the first predetermined value, the zero point correction unit 513 determines whether the first predetermined number of data (the first predetermined number of data The zero point of the sensor may be corrected using various calculation results using data. Since the zero point correction processing by the zero point correction unit 513 is known, the detailed description will be omitted.

  Next, with reference to FIG. 2, the first zero point correction process by the ECU 50 of the embodiment will be described. FIG. 2 is a flowchart showing the procedure of a first zero point correction process by the ECU 50 of the embodiment. In the following, among the operations of the processing unit 510, when the operations other than the acquisition unit 511, the determination unit 512, and the zero point correction unit 513 are described, the operation subject is referred to as the processing unit 510.

  First, in step S1, the processing unit 510 executes initialization (reset) of the RAM. Next, in step S2, the acquisition unit 511 acquires one sensor data (sample data). Next, in step S3, processing unit 510 determines whether or not two or more pieces of data have been acquired, and in the case of Yes, the process proceeds to step S4, and in the case of No, the process returns to step S2.

  In step S4, the determination unit 512 determines whether or not the degree of variation of the acquired two or more data is less than a predetermined value. If the determination is YES, the process proceeds to step S5. If the determination is NO, the process returns to step S1. In step S4, different determination methods or different determination values can be taken depending on whether the number of data is a predetermined number (for example, 100) or less. The details of the process of step S4 will be described later.

  In step S5, processing unit 510 determines whether or not a predetermined number (for example, 100) of data has been acquired. If the determination is YES, the process proceeds to step S6, and if the determination is NO, the process returns to step S2. In step S6, the zero point correction unit 513 corrects the zero point of the sensor using various calculation results of a predetermined number of data.

  Next, with reference to FIG. 3, an outline of the entire processing in the embodiment will be described. FIG. 3 is a graph showing the relationship between the calculated value of data and the threshold value E in the embodiment. Here, N is the number (first predetermined number) of data to be acquired in order to perform the zero point correction process of the sensor. Also, S (second predetermined number) is a number smaller than N. Further, in the graph of FIG. 3, the vertical axis represents a calculation value when a predetermined calculation (for example, calculation by the RMS method described later) is performed on data. E is a threshold (first predetermined value). The code L1 represents the transition of the operation value. Note that this calculated value does not decrease (details will be described later). The horizontal axis represents time, and "N" and "S" in the horizontal axis are "time required to acquire N pieces of data" and "time required to acquire S pieces of data, respectively. Represents ".

  As in the prior art, also in the present embodiment, if the calculated value by the acquired N data is less than the predetermined value E, it is determined that the N data is normal, and the sensor is used using the N data Perform zero correction of. The present embodiment differs from the prior art in that, for example, when S pieces of data are obtained, as shown by a code L2, when the calculated value becomes a predetermined value E or more, the determination is reset at that time. (The details will be described later).

  Next, specific calculation contents will be described. When zero point correction is performed using a predetermined number of data (sample data), it is important that all the predetermined number of data is normal data, that is, it is not abnormal data affected by large noise etc. . In that case, for example, when all the values of the predetermined number of data are within the predetermined range, it can be determined that all are normal data. Hereinafter, this determination method is referred to as “predetermined range determination method”. In addition, since the amplitude of noise fluctuates according to a certain probability distribution, the probability of occurrence of data affected by a large noise increases as the predetermined number increases.

  And in the above-mentioned predetermined range judgment method, since the difference between the currently set zero value and the actual zero value can not be taken into consideration, even if the predetermined number of data is actually all normal data Data outside the range may be determined as abnormal data.

  One of the methods that can solve this shortcoming of the predetermined range determination method is an RMS (Root Mean Square) method. In the RMS method, only the degree of data variation can be determined without setting a predetermined range. Specifically, the degree of data variation can be determined using the following equations (1) and (2).

ここで、Ｎは所定数である。ｘ_ｉはｉ番目（１≦ｉ≦Ｎ）のデータの値である。Ｅは閾値である。

Here, N is a predetermined number. x _i is the value of the ith (1 ≦ i ≦ N) data. E is a threshold.

  In this RMS1 method, N pieces of data are stored in a buffer (storage unit), then RMS1 is calculated, and it is determined whether the RMS1 is less than a predetermined value E or not. If RMS1 is less than the predetermined value E, the N pieces of data are used to perform the zero point correction. If RMS1 is equal to or greater than the predetermined value E, the acquired data is once reset, for example, to discard the acquired data, and data acquisition and the determination according to equation (1) are repeated again until RMS1 becomes less than the predetermined value E. However, according to this RMS1 method, a buffer capable of storing all of N pieces of data is required, leading to an increase in cost.

Therefore, as a method of solving the problem of the buffer, there is a method of RMS2 which is determined using the following equation (3) and the above equation (2).

  In this RMS2 method, a buffer for storing the first term in the root of equation (3) and a buffer for storing the second term are prepared, and the first term and the second term are updated each time data is acquired. Then, RMS2 is calculated, and it is determined whether the RMS2 is less than a predetermined value E or not. If RMS2 is less than the predetermined value E, N pieces of data can be used for the zero point correction process, and the zero point correction is performed using various calculation results. If RMS2 is equal to or greater than the predetermined value E, reset is performed, and a predetermined number of data are obtained again, and the determination by the equation (3) is performed. According to this RMS2 method, the required buffer capacity can be reduced, but the determination can be made only after acquiring a predetermined number of data, and the improvement can be made in that the zero point correction can not be completed early. There is room.

So, in this embodiment, the method of RMS3 by the following formula (4) is adopted.

Here, N is a first predetermined number. x _i is the value of the ith (1 ≦ i ≦ N) data. E is a first predetermined value (threshold). S (2 ≦ S <N) is a second predetermined number. This RMS3 is a root mean square assuming that the “S + 1” th to Nth data are the variance 0 (the deviation from the average value of the S data is 0) when the Sth sample data is acquired. . Therefore, RMS3 is an increasing function that never decreases. That is, if RMS3 becomes equal to or greater than the predetermined value E when the S-th data x _s of the data x _i is acquired, then the RMS 3 (equivalent to RMS 2) when N pieces of data are aligned is also the predetermined value E The above is finalized. That is, the second predetermined value described above is 100%.

  The determination unit 512 performs a reset process when the above equation (4) is not satisfied. That is, the determination unit 512 performs the reset when the RMS 3 becomes equal to or more than the predetermined value E based on the S pieces of data. In other words, the determination unit 512 determines that the first predetermined number is N, the second predetermined number is S, S pieces of data, and N−S pieces of data assuming that the variation is 0 (zero). Reset processing is performed based on the result of the degree of variation calculated based on the above.

  Next, with reference to FIG. 4, a flow of zero point correction processing in the case of using the method of RMS2 and in the case of using the method of RMS3 will be described. FIG. 4 is a graph showing a flow of zero point correction processing when the method of RMS2 and the method of RMS3 are used in the embodiment.

In FIG. 4, it is assumed that sample data of values as illustrated are sequentially acquired. Then, the sample data x ₃ is the data affected by large noise or the like. When the method of RMS 2 is used, the calculation result of the sample data (x _{1 to} x _s ) in the middle can be large or small depending on the subsequent sample data (x _{S + 1 to} x _N ). if not x ₁ ~x _N) calculation results using, variations in the sample data _(x 1 ~x _N) can not be reliably determined whether less than a predetermined value E. Therefore, after the sample data x _N is obtained, the determination is made according to the equation (3), and if the determination result is NG, the sample data is reset and N sample data (x _{N + 1 to} x _{2 N} ) are obtained again. If the determination result is OK, zero point correction is performed thereafter to complete learning (zero point correction processing).

On the other hand, in the case of using the method of RMS3, after obtaining the _second sample data (x ₂ ), the determination by the equation (4) is performed each time the sample data is obtained. Then, to obtain the sample data x ₃ was determined according to formula (4) from the determination result is reset because it is NG. Thereafter, sample data is acquired again, the determination by equation (4) is repeated, N pieces of sample data are acquired, and if the determination result by equation (4) is OK, zero point correction is performed thereafter, Learning (zero point correction processing) can be completed. That is, when the method of RMS3 which is an increasing function that does not decrease is used, when sample data x _{N + 3} is acquired much faster than when sample data x _{2 N} is acquired when RMS 2 method is used. , Can finish learning.

  FIG. 5 illustrates the flow of the zero point correction process in the case where the yaw rate sensor value is acquired and the method of RMS2 is used and the method of RMS3 is used in the embodiment. FIG. 5 is a graph showing the flow of the zero point correction process when the yaw rate sensor value is acquired and the method of RMS2 is used and the method of RMS3 is used in the embodiment.

  When using a yaw rate sensor value (value of detection data by the yaw rate sensor SE9), for example, the yaw rate sensor value at the time of straight running of the vehicle (when the steering angle is approximately 0 and the difference between the left and right wheel speeds is approximately 0) The yaw rate zero point is calculated from the average value of the data by sampling a predetermined number of times. At that time, if the acquired data is equal to or more than the predetermined variation degree, it is resampled as unreliable.

  When the required number of sample data increases, the probability that the data affected by noise etc. is included in the sample data increases, the degree of variation becomes more than a predetermined level, and sampling is often repeated, and the zero point correction process is performed. It will be difficult to complete.

  For example, when correcting the zero point using ten pieces of data, as shown in FIG. 5, data No. 1 to No. 33 of data are acquired, and data No. 4 and No. 19 in the data are obtained. It is assumed that No. 20 is mixed with data affected by noise. In that case, in the RMS2 method, the first sampling (data No. 1 to No. 10) is judged as NG and reset, and the second sampling (data No. 11 to No. 20) is judged as NG. The determination is made and reset, and the third sampling (data No. 21 to No. 30) is made, and it is determined that the determination is OK for the first time, and then the zero point correction processing can be completed.

  On the other hand, in the method of RMS3, the sampling of data No. 4 is sampled at the time of sampling first, then it is judged as NG and reset, and data No. 5 to No. 14 are acquired by resampling. It can be determined as OK, and the zero point correction process can be completed thereafter. That is, according to the method of RMS3, it is possible to re-sample data earlier and complete the zero point correction processing as compared with the method of RMS2.

Next, another determination criterion will be described. As described above, when RMS3 becomes equal to or larger than the predetermined value E when the S-th data x _S of the data x _i is acquired, the variation of N pieces of sample data when N pieces of data are aligned thereafter The probability that the degree was equal to or higher than the predetermined value E was 100%. However, since it is unlikely that the “S + 1” th to Nth data become the variation 0 (the deviation from the average value of the S data is 0), the probability is a value slightly smaller than 100% (for example, 99 It is also effective to use%, 95%, etc.).

In the following, it is assumed that the values of sample data follow a normal distribution. Then, based on the S sample data, it is determined whether the index value Z is less than the threshold value E4 (third predetermined value), as shown in the following equation (5).

  In other words, determination unit 512 determines that the second predetermined number is S, and the probability that the variation degree of the first predetermined number of data is equal to or greater than the first predetermined value is second based on S pieces of data. A determination value for determining whether or not to be a predetermined value or more is calculated based on the chi-square distribution of the degree of freedom S-1.

  Here, FIG. 6 is a graph showing the relationship between the index value Z and the threshold value E4 in the embodiment. The vertical axis represents the value of the index value Z. The code L11 represents the transition of the index value Z. The horizontal axis represents time, and "N" and "S" in the horizontal axis are "time required to acquire N pieces of data" and "time required to acquire S pieces of data, respectively. Represents ". When the index value Z calculated by acquiring the sample data up to the S-th becomes E4 or more (symbol L12), then, when acquiring the sample data up to the N-th, the variation degree of the N sample data Is greater than or equal to the predetermined value E indicates that the probability is greater than or equal to the second predetermined value. Hereinafter, how to determine the threshold value E4 will be described.

  Here, FIG. 7A is a graph showing a standard normal distribution (mean is 0, variance is 1). FIG. 7 (b) is a graph showing a normal distribution. FIG. 7C is a graph showing the cumulative distribution function of the chi-square distribution corresponding to the standard normal distribution. FIG. 7D is a graph showing the probability density function of the chi-square distribution corresponding to the normal distribution.

  Assuming that the sample data follows the standard normal distribution, the index value Z calculated with S sample data follows the chi-square distribution with “S−1” degrees of freedom. Therefore, for the standard normal distribution in FIG. 7A, the numerical value E2 is such that the cumulative probability of the chi-square distribution with the degree of freedom “S−1” is S1 (eg, 0.95) when the number of sample data is S. calculate. At that time, the probability that the index value Z becomes equal to or less than E2 is P1.

  Then, when the sample data follows the normal distribution (FIG. 7 (b)), since it can be interconverted with the standard normal distribution (FIG. 7 (a)), E2 in the case of the standard normal distribution is the normal distribution of the sample data. Convert to the case value to obtain E4 (FIG. 7 (d)). Thus, E4 can be calculated. Then, when the S sample data are acquired, if the index value Z becomes E4 or more, it may be determined as NG and reset.

  Further, E4 in FIG. 6 may be approximated to a straight line or a broken line to be E41. At this time, a least squares method or the like may be used to form approximate straight lines or broken lines.

  Also, instead of calculating the numerical value E2 in which the number of sample data is S and the cumulative probability of the chi-square distribution with the degree of freedom “S-1” has a predetermined probability P1 (for example, 0.95), the number of sample data is N In consideration of the probability that RMS3 will be equal to or greater than the predetermined value E when this becomes, calculate the numerical value E5 that the cumulative probability of the chi-square distribution with the degree of freedom “NS−1” becomes the predetermined probability P2 (for example 0.05) Alternatively, E5 in the case of the standard normal distribution may be converted to a value in the case of the normal distribution of sample data to be E4. In this case, when the index value Z becomes E4 or more, it becomes NG with the probability of 1-P2 (for example, 0.95 when P2 is 0.05), so when S sample data are obtained If the index value Z becomes E4 or more, it may be determined as NG and reset. E4 in that case may also be an approximate straight line or a broken line using the least squares method or the like.

Further, as another threshold, E6 (using only the portion of E6 <E) may be used instead of E as a threshold for RMS3. E6 can be expressed as follows using, for example, a predetermined value E7.
E6 = E7 × √ (S / N)

  By using this E6, NG can be determined and reset earlier than in the case of using E. E7 can be determined by the value of variance of noise in the specification of the sensor. Moreover, it is preferable to set it as E7> E.

  Next, the second zero point correction process by the ECU 50 of the embodiment will be described with reference to FIG. FIG. 8 is a flowchart showing the procedure of a second zero point correction process by the ECU 50 of the embodiment.

  First, in step S11, processing unit 510 executes initialization (reset) of the RAM. Next, in step S12, the acquisition unit 511 acquires one sensor data (sample data). Next, in step S13, processing unit 510 determines whether or not two or more pieces of data have been acquired, and in the case of Yes, the process proceeds to step S14, and in the case of No, the process returns to step S12.

  In step S14, the determination unit 512 calculates the average value of the data acquired so far. Next, in step S15, the determination unit 512 determines whether or not the degree of deviation from the average value is less than a predetermined value (fourth predetermined value) for the data acquired so far, and in the case of Yes, step S18. If No, the process proceeds to step S16.

  In step S16, the determination unit 512 discards data whose deviation degree from the average value is equal to or more than a predetermined value. Next, in step S17, the determination unit 512 determines whether the number of times of data discarding is equal to or greater than a predetermined number. If yes, the process returns to step S11, and if no, the process proceeds to step S18. When the process returns to step S11, the determination unit 512 performs a reset process of selecting new data without using all the data acquired so far.

  In step S18, processing unit 510 determines whether or not a predetermined number (eg, 100) of data has been acquired. If the determination is yes, the process proceeds to step S19, and if the determination is No, the process returns to step S12. In step S19, the zero point correction unit 513 corrects the zero point of the sensor using various calculation results of a predetermined number of data.

  The process of FIG. 2 and the process of FIG. 8 may be performed by either one or both of them may be performed in parallel.

  In this manner, according to the ECU 50 of the present embodiment, the process performed based on the plurality of acquired data can be completed early. Specifically, for example, before determining the degree of variation of N pieces of data, the degree of variation of N pieces of data can be predicted with high accuracy based on S pieces of data less than N pieces of data. A reset can be implemented, thereby achieving an early zero correction completion.

  More specifically, for example, based on S pieces of data using RMS 3 (Equation (4)), the variation degree of N pieces of data thereafter becomes the predetermined value E or more with a probability of 100%. It can be determined.

  Also, for example, using index value Z (Equation (5)), based on the S data, it may be determined that the degree of variation of the N data thereafter becomes the predetermined value E or more with high probability. it can.

  Further, according to the process of FIG. 8, the influence of noise can be suppressed by discarding data having a large deviation from the average value or resetting when the number of discards is equal to or more than a predetermined number.

  Further, unlike the prior art, the zero point that can be learned is not limited to only within a predetermined range, so zero point learning of an arbitrary size can be realized. That is, it is possible to cope with the case where the current zero point is far from the initial value or the previous value.

  Further, since it is not necessary to store all sample data in the buffer, the capacity of the buffer can be reduced even if the number of data is large.

  Further, in particular, by using RMS 3 (formula (4)), the following effects can be obtained. In the above-mentioned conventional method using RMS1 (as well as RMS2), even if the value of RMS1 becomes larger than the threshold when sample data is acquired halfway, the final value of RMS1 becomes larger than the threshold. Not exclusively. On the other hand, it is possible to predict that the final value of RMS1 exceeds the threshold with a probability of 100% or a high probability by using RMS3 (Equation (4)) which is an increasing function that does not decrease.

  As mentioned above, although the embodiment of the present invention was illustrated, the above-mentioned embodiment is an example to the last, and limiting the scope of the present invention is not intended. The above embodiment can be implemented in other various forms, and various omissions, replacements, combinations, and changes can be made without departing from the scope of the invention. In addition, the specifications (structure, type, number, etc.) of each configuration, shape, etc. can be appropriately changed and implemented.

  For example, the sensor to which the zero point correction process of the present embodiment is applied is not limited to the yaw rate sensor, and may be another sensor such as an acceleration sensor.

  In addition, other than the zero point correction process, for example, the actual pressure at the time of performing a predetermined output by the pressing mechanism is acquired, and the process is performed based on a plurality of sample data, such as when determining from a plurality of sample data. It can be widely applied to

  11 ... driving force generating device, 12 ... steering device, 13 ... braking device, 50 ... ECU, 510 ... processing unit, 511 ... acquisition unit, 512 ... determination unit, 513 ... zero point correction unit (data processing unit), 520 ... Memory unit, FR, FL, RR, RL ... wheels.

Claims (8)

An acquisition unit for acquiring data,
The second predetermined number of data less than the first predetermined number by the acquisition unit before determining whether the degree of variation of the first predetermined number of data is less than the first predetermined value. The probability that the variation degree of the first predetermined number of data becomes equal to or higher than the first predetermined value is equal to or higher than the second predetermined value based on the second predetermined number of data when If it is determined that the probability is equal to or higher than the second predetermined value, the reset process is performed to select new data without using all or part of the second predetermined number of data. A determination unit to perform
Data processing for performing a predetermined process using the first predetermined number of data when it is determined by the determination unit that the variation degree of the first predetermined number of data is less than the first predetermined value And an electronic device of a vehicle. The determination unit is
The variation calculated based on the first predetermined number is N, the second predetermined number is S, and the S pieces of data and N−S pieces of data assuming that the variation is 0 (zero). The vehicle electronic device according to claim 1, wherein the reset process is performed based on a degree result. The first predetermined number is N, the data is x _i (1 ≦ i ≦ N), the first predetermined value is E, the second predetermined number is S (2 ≦ S <N), the second Let 100% be the predetermined value of
The electronic device of the vehicle according to claim 2, wherein the determination unit performs the reset process when the following expression (11) is not satisfied.

The determination unit is
The second predetermined number is S, and the probability that the variation degree of the first predetermined number of data becomes equal to or higher than a first predetermined value is equal to or higher than a second predetermined value based on the S pieces of data. The electronic device of the vehicle according to claim 1, wherein a determination value for determining whether or not to be calculated is calculated based on a chi-square distribution of degrees of freedom S-1. The first predetermined number is N, the data is x _i (1 ≦ i ≦ N), the first predetermined value is E, the second predetermined number is S (2 ≦ S <N), the second If the third predetermined value corresponding to the predetermined value of
The electronic device of the vehicle according to claim 1, wherein the determination unit performs the reset process when the following expression (12) is not satisfied.

  The determination unit calculates an average value of the data acquired by then after the second predetermined number of data is acquired, and a deviation degree from the average value is equal to or more than a fourth predetermined value. The vehicle electronic device according to claim 1, wherein the data is discarded.   The electronic device of a vehicle according to claim 6, wherein, when the number of discards is equal to or more than a predetermined number, the determination unit performs a reset process of selecting new data without using all of the second predetermined number of data. . The acquisition unit acquires data detected by a sensor,
The data processing unit performs zero point correction that corrects a zero point that is a reference value of the sensor, using the first predetermined number of data. Electronic device of the described vehicle.

Images