JP2015190920A - Measurement device, and failure diagnosis device using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately calibrate the scale factor error and off-set error of a yaw rate detected by a yaw rate sensor.SOLUTION: A calibration yaw rate computing unit 44 prepares coordinate data while the yaw rate γ detected by a yaw rate sensor 34 and the lane estimated yaw rate γlane estimated from a lane compartment line recognized by image recognition are used as coordinate axes, and determines a linear regression parameter by the least square method on the basis of the image data. The computing unit 44 identifies the gradient and intercept of the linear regression parameter as the scale factor error and off-set error, and determines a calibrated yaw rate γcal by calibrating the yaw rate γ using both errors.

Description

  The present invention has a physical quantity measuring means for directly measuring a physical quantity and a physical quantity estimating means for estimating the physical quantity, and is based on the first physical quantity measured by the physical quantity measuring means and the second physical quantity estimated by the physical quantity estimating means. The present invention relates to a measuring device for obtaining a calibrated physical quantity obtained by calibrating a first physical quantity, and a failure diagnosis apparatus using the measuring device.

  Conventionally, measurement values output from a measurement device such as a yaw rate sensor mounted on a vehicle are likely to cause errors due to the influence of heat damage, etc. Therefore, the output measurement values are neutrally calibrated (zero point calibration). Various techniques have been proposed. For example, in Patent Document 1 (Japanese Patent No. 5314649), a value measured by a reference measurement device is compared with a value measured by a measurement device different from the measurement device, and is constantly measured by moving average. A technique for performing neutral calibration using offset components as offset components is disclosed.

  Further, for example, in Patent Document 2 (Japanese Patent Application Laid-Open No. 2011-169728), when performing neutral calibration of the yaw rate sensor, the yaw rate is estimated from the lane recognized based on the image captured by the camera, and based on this estimated value. A technique for calibrating the yaw rate is disclosed.

  Furthermore, in Patent Document 3 (Japanese Patent No. 3637801) and Patent Document 4 (Japanese Patent No. 4432885), a yaw rate is estimated based on a steering angle signal detected by a steering angle sensor, and based on this estimated yaw rate. A technique for detecting a failure of a yaw rate sensor is disclosed.

Patent No. 5314649 JP 2011-169728 A Japanese Patent No. 3637801 Japanese Patent No. 4342885

  By the way, when the vehicle is going to travel straight in a state where a crossing slope (cant) has a crossing slope (cant) or in a state of receiving a crosswind, the driver travels with the steering wheel turned off. In such a situation, as disclosed in Patent Document 1, when neutral calibration of the yaw rate sensor is performed based on the steering angle, the above-described road crossing gradient and cross wind act as a lateral force disturbance, resulting in a large error. Therefore, there is a disadvantage that the neutral calibration accuracy of a measuring device such as a yaw rate sensor is lowered.

  In addition, since the yaw rate estimated from the steering angle has a phase characteristic different from that of the actual vehicle yaw rate, it is necessary to take measures to limit to a steady state or to perform a phase matching process in order to exclude the phase characteristic error. . In this case, the correction frequency is reduced, and accurate estimation / correction becomes difficult due to errors in the phase alignment itself.

  In this regard, as disclosed in Patent Document 2, a technique for detecting a lane based on an image captured by a camera and estimating a yaw rate based on the detected lane is hardly affected by a road surface gradient or a crosswind. Compared with the technique disclosed in the cited document 1, neutral calibration can be performed with high accuracy. However, there is an inconvenience that an error due to the difference in phase characteristics between the two is included due to a delay in image processing and the like as in the means described in Patent Document 1.

  In addition, a measurement apparatus using a rate gyro such as a yaw rate sensor has a scale factor error as an error factor in addition to an offset error. In the technique disclosed in Patent Document 2, since the offset error is detected in a state in which the scale factor error factor is included, the scale factor error is estimated as an offset error, and vice versa. There is a disadvantage that the error estimation and calibration accuracy cannot be increased.

  Further, as disclosed in Patent Documents 3 and 4, when a steering angle sensor is used as a failure diagnosis of a measuring device, it is easily affected by a road surface crossing gradient or a cross wind as described above, and the failure diagnosis is performed with high accuracy. There is a bug that cannot be done.

  In view of the above circumstances, the present invention can individually detect an offset error and a scale factor error, is not easily affected by the phase characteristics, and performs calibration of the scale factor error and the offset error with high accuracy. An object of the present invention is to provide a measuring device capable of performing the above and a failure diagnosis device using the same.

  The measuring apparatus according to the present invention includes a physical quantity measuring unit that measures a first physical quantity, a physical quantity estimating unit that sets a second physical quantity that estimates the first physical quantity, and that is different from the physical quantity measuring unit, and the first physical quantity Coordinate data acquisition means for acquiring coordinate data using the second physical quantity as a coordinate axis, parameter calculation means for obtaining a linear regression parameter by the least square method based on the acquired coordinate data, and the slope and intercept of the linear regression parameter Parameter identifying means for identifying as a scale factor error and an offset error; and a physical quantity calibration means for calibrating the first physical quantity measured by the physical quantity measuring means based on the scale factor error and the offset error to obtain a calibration physical quantity. Prepare.

  The failure diagnosis device using the measuring device according to the present invention compares the scale factor error and the offset error with a preset failure determination threshold, and at least one of the errors exceeds the failure determination threshold. If there is, a failure diagnosis means for determining that the physical quantity measurement means is faulty is provided.

  According to the present invention, the first physical quantity and the second physical quantity for estimating the first physical quantity are obtained by different devices, and the linear regression parameters are obtained by the least square method based on the coordinate data obtained using the obtained physical quantities as coordinate axes. Since the slope and intercept of the linear regression parameter are identified as the scale factor error and the offset error and the first physical quantity is calibrated, the physical quantity obtained by the calibration is separately calibrated by the offset error and the scale factor error. Therefore, the scale factor error and the offset error can be calibrated with high accuracy without being affected by the phase characteristics.

Schematic configuration diagram of vehicle steering system Functional block diagram of the steering control unit Flowchart showing the yaw rate correction value calculation routine (part 1) Flowchart showing the yaw rate correction value calculation routine (part 2) Explanatory drawing which shows the procedure which estimates a yaw rate based on the curve approximation of a curve road Explanatory diagram when going straight on a road with crossing slope (A) is explanatory drawing of the state which plotted coordinate data on the coordinate, (b) is a histogram of a yaw rate, (c) is explanatory drawing which shows dispersion | distribution of a histogram. Explanatory drawing of coordinates with the yaw rate on the vertical axis and the estimated yaw rate on the horizontal axis

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Reference numeral 1 in FIG. 1 denotes an electric power steering apparatus in which the steering angle can be set independently of the driver input. In this electric power steering apparatus 1, a steering shaft 2 is rotatably supported by a body frame (not shown) via a steering column 3, one end of which extends to the driver's seat side, and the other end of the engine room side. It is extended to. A steering wheel 4 is fixed to an end portion of the steering shaft 2 on the driver's seat side, and a pinion shaft 5 is connected to an end portion extending to the engine room side.

  A steering gear box 6 extending in the vehicle width direction is disposed in the engine room, and a rack shaft 7 is inserted into and supported by the steering gear box 6 so as to be reciprocally movable. A rack (not shown) formed on the rack shaft 7 is engaged with a pinion formed on the pinion shaft 5 to form a rack and pinion type steering gear mechanism.

  The left and right ends of the rack shaft 7 protrude from the end of the steering gear box 6, and a front knuckle 9 is connected to the end via a tie rod 8. The front knuckle 9 rotatably supports left and right wheels 10L and 10R as steering wheels and is supported by a vehicle body frame so as to be steerable. Accordingly, when the steering wheel 4 is operated and the steering shaft 2 and the pinion shaft 5 are rotated, the rack shaft 7 is moved in the left-right direction by the rotation of the pinion shaft 5, and the front knuckle 9 is moved by the movement to the kingpin shaft (not shown). And the left and right wheels 10L, 10R are steered in the left-right direction.

  An electric power steering motor (electric motor) 12 is connected to the pinion shaft 5 via an assist transmission mechanism 11, and assists and sets steering torque applied to the steering wheel 4 by the electric motor 12. The target yaw moment (steering torque) is added so that the yaw rate (target yaw rate) is reached. The electric motor 12 is driven by the driving force output from the motor driving unit 21. A control signal corresponding to the target steering torque is input to the motor drive unit 21 from a steering control unit 20 described later.

  On the input side of the steering control unit 20, lane markings represented by the left and right white lines that regulate the driving lane are detected, and as front recognition means and lane marking calculation means for acquiring position information of these lane markings A vehicle speed sensor 32 for detecting a vehicle speed V, a steering angle sensor 33 for detecting a steering angle, a yaw rate sensor 34 as a physical quantity measuring means for detecting a yaw rate γ which is a first physical quantity, and the like are connected. Based on these parameters, electric power steering control that assists the driver's steering force, lane keeping control that causes the vehicle to travel along the target traveling path, lane departure prevention control that prevents lane departure from the traveling lane, and the like are executed. . In the following, the lane departure prevention control function of each control will be described.

  The front recognition device 31 is, for example, a stereo camera that is attached to the front of a ceiling in a vehicle interior at a fixed interval and that includes a main camera and a sub camera that stereo-shoots an object outside the vehicle from different viewpoints, and image data from both cameras. And a stereo image processing unit for processing.

  Further, the processing of image data from each camera in the stereo image processing device of the forward recognition device 31 is performed as follows, for example. First, distance information is obtained from the parallax between the reference image captured by the main camera and the comparison image captured by the sub camera, and a distance image is generated. Based on the knowledge that the lane markings that regulate the left and right of the driving lane are brighter than the road surface, the luminance change in the width direction of the road is evaluated, and the positions of the left and right lane markings on the image plane are determined. Specify on the image plane.

  The biaxial coordinates (x, z) of the lane marking on the virtual plane are calculated by a known coordinate conversion formula based on the position (i, j) on the image plane and the distance information calculated with respect to this position. The In the present embodiment, for example, as shown in FIG. 5, the coordinate system of the real space set based on the position of the host vehicle is based on the road surface directly below the center of the stereo camera, and the vehicle width direction is the X axis ( The left direction is “+”) and the vehicle length direction (distance direction) is the Z-axis (front direction is “+”). At this time, the XZ plane coincides with the road surface when the road is flat.

  The road model is expressed by dividing the traveling lane of the host vehicle on the road into a plurality of sections in the distance direction, and connecting the left and right lane markings in each section to a predetermined approximation. In the present embodiment, the example of recognizing the shape of the driving lane based on images from a set of cameras has been described. However, the shape is obtained based on image information from a monocular camera and a color camera. Also good. Further, as long as the distance to the lane marking can be recognized, the forward recognition device 31 may be a millimeter wave radar or an infrared laser radar.

  Next, the lane departure prevention control executed by the steering control unit 20 will be described. As shown in FIG. 2, the steering control unit 20 has functions for executing lane departure prevention control, such as a target yaw rate calculation unit 41, a feedforward control value calculation unit 42, an estimated yaw rate calculation unit 43 as physical quantity estimation means, and a physical quantity calibration. A calibration yaw rate calculator 44, a feedback control value calculator 45, a rate limiter calculator 46, and a steering torque calculator 47 are provided as means.

  The target yaw rate calculation unit 41 calculates the target yaw rate γ * as follows. Note that the lane yaw angle calculation means of the present invention is included in the target yaw rate calculation unit 41.

  That is, as shown in FIG. 5, the left and right lane line candidate points PL and PR calculated based on the distance data are plotted on the virtual road plane generated based on the reference image data and the comparison image data. The host vehicle M is disposed on the virtual road plane, and a point on the road surface immediately below the center of the front recognition device (stereo camera) 31 mounted on the host vehicle M is used as an origin, and the host vehicle M Coordinates that take the x-axis in the vehicle width direction and the z-axis in the front-rear direction are set. The symbol θyaw is an angle formed by the traveling lane and the host vehicle M (to-lane yaw angle).

Then, from the group of candidate points PL and PR plotted on the left and right, the quadratic model formula of the left and right lane markings is obtained by the least square method. Y_LINE_L = A_L · Z ² + B_L · Z + C_L
Y_LINE_R = A_R · Z ² + B_R · Z + C_R (2)
The coefficients A_L, B_L, C_L, A_R, B_R, and C_R are obtained, thereby obtaining the left and right lane marking models. Here, A_L and A_R are parameters relating to the lane curvature, the curvature κ of the left lane marking is κ = 2 · A_L, and the curvature κ of the right lane marking is κ = 2 · A_R. B_L and B_R are angle parameters related to the inclination (yaw angle) of the host vehicle M with respect to a virtual line passing through the center of the host vehicle M. Further, C_L and C_R are coefficients indicating the lateral position from the left and right lane markings.

And the estimated lane departure time Tttlc of the vehicle M is
θyaw = (B_L + B_R) / 2 (3)
Ask from.

Also, the vehicle lateral position Y in the lane width direction, which is the vehicle position from the center of the traveling lane,
Y = (C_L + C_R) / 2 (4)
Ask from.

Next, the lane line _inter-vehicle distance L
L = ((C_L-C_R) -TR) / 2-Y (5)
Calculate from Here, TR is a tread of the vehicle, and in the present embodiment, the outer end position of the tire is used as a reference for the lane departure determination.

Then, the estimated lane departure time Tttlc deviating from the driving lane is, for example,
Tttlc = L / (V · sin (θyaw)) (6)
Ask from.

Based on the anti-lane yaw angle θyaw and the estimated lane departure time Tttlc, the target yaw rate γ * for preventing the lane departure is
γ * = − θyaw / Tttlc (7)
Ask from.

 Then, the feedforward control value calculation unit 42 corrects the target yaw rate γ * with a predetermined gain and sets the feedforward value Mzff.

On the other hand, the estimated yaw rate calculation unit 43 is based on the lane curvature κ and the lane yaw angle θyaw of the lane line determined based on the image recognized by the forward recognition device 31 and the vehicle speed V detected by the vehicle speed sensor 32. 2 Estimated yaw rate γlane, which is a physical quantity,
γlane = κ · V + dθyaw / dt (8)
Ask from.

  Since the estimated yaw rate γlane is obtained based on the image recognized by the forward recognition device 31, the yaw rate can be estimated with high accuracy without being affected by the road surface gradient (kant) or the crosswind. That is, as shown in FIG. 6, when the host vehicle M travels on a straight road having a road surface crossing gradient, a downward lateral force Fy acts on the vehicle body, so the driver turns the steering wheel upward. Drive straight in the state. Therefore, when the yaw rate is estimated based on the steering wheel angle, the lateral force Fy acts as a disturbance, so that the yaw rate cannot be estimated with high accuracy.

  On the other hand, in this embodiment, since the estimated yaw rate γlane is obtained based on the curve approximation formula of the left and right lane markings obtained from the image recognized by the front recognition device 31, the lateral force Fy is applied to the host vehicle M. Even if it acts, since it does not act as a disturbance when obtaining the lane marking, the estimated yaw rate γlane can be obtained with high accuracy.

  Then, the calibration yaw rate calculation unit 44 obtains a calibration yaw rate γcal obtained by calibrating the yaw rate γ from the relationship between the estimated yaw rate γlane and the yaw rate γ detected by the yaw rate sensor 34. Specifically, the processing in the calibration yaw rate calculation unit 44 is executed according to a calibration yaw rate calculation routine shown in FIG.

  In this routine, first, in step S1, an estimated yaw angular acceleration dγlane / dt is obtained by time differentiation of the estimated yaw rate γlane.

  Next, the process proceeds to step S2, and it is checked whether or not the estimated yaw angular acceleration dγlane / dt is greater than or equal to a preset threshold value γsl. The estimated yaw angular acceleration dγlane / dt is sample data for performing the least squares described later, and the threshold γsl is a value preferable as sample data by excluding those with small fluctuations in the estimated yaw angular acceleration dγlane / dt. Is set to extract.

  If dγlane / dt ≧ γsl, the process proceeds to step S3, and sampling of the estimated yaw rate γlane and the yaw rate γ detected by the yaw rate sensor 34 synchronized therewith (hereinafter referred to as “sensor yaw rate γsensor”) is started. . On the other hand, if dγlane / dt <γsl, the routine is exited.

  In step S3, it is checked whether the sample data number n of the estimated yaw rate γlane is less than a preset threshold value nsl. This threshold value nsl is a value for determining whether or not the number n of sample data sufficient as data for performing the least-squares described later is reached. In this embodiment, the threshold value nsl is based on the memory capacity of the steering control unit 20. Is set. If the memory capacity is sufficient, the threshold value nsl may be obtained and set in advance from simulation or the like.

  If n <nsl, the process proceeds to step S5. If n ≧ nsl, the process branches to step S4. In step S4, coordinate data for least squares is created and acquired based on the sampled estimated yaw rate γlane and sensor yaw rate γsensor. Note that the processing in this step corresponds to the coordinate data acquisition means of the present invention.

  That is, as shown in FIG. 7A, a function f (γlane, γsensor), which is coordinate data, is plotted on a coordinate plane in which the value of the estimated yaw rate γlane is taken on the abscissa axis and the value of the sensor yaw rate γsensor is taken on the ordinate axis. To do. By the way, when the sample data is dispersed on average, a linear regression equation (γsensor = a · γlane + b) can be obtained with high accuracy.

  For this reason, in step S4, a histogram of a predetermined floor indicating the frequency distribution of the sensor yaw rate γsensor is created on the estimated yaw rate γlane axis as shown in FIG. Then, as shown in FIG. 6C, the sensor yaw rate γsensor of the floor showing the frequency above a certain threshold SL is moved from the floor showing the maximum frequency to the floor showing the next frequency, Average. Thereafter, when the least square data is created, the process jumps to step S7.

  On the other hand, when it is determined in step S3 that the number of sample data n is less than the predetermined threshold value nsl and the process proceeds to step S5, it is checked whether or not the value of the sensor yaw rate γsensor is zero. When γsensor = 0, since the value exceeds the resolution of the yaw rate sensor 34, the process jumps to step S7 without data sampling. On the other hand, if γsensor ≠ 0, the process proceeds to step S6.

  In step S6, a process for storing the sample data read this time is executed. That is, the number n of sampled data (γlane, γsensor) is sequentially stored until the threshold value nsl set according to the memory capacity is reached.

  And step S4, step S5. Or if it progresses to step S7 from step S6, the process which calculates a yaw rate detection range will be performed. This yaw rate detection range is between the maximum value and the minimum value described above, and the variance of the sampled data (γ lane, γ sensor) is checked. Even if the number of samples of the data (γlane, γsensor) stored in the memory in step S6 does not reach the memory threshold value, the detection range (maximum value−minimum value) of the yaw rate is calculated in step S7. .

  Thereafter, when the process proceeds to step S8, it is checked whether or not the number n of the sampled least square data (γlane, γsensor) is equal to or greater than the threshold value nsl. If n ≧ nsl, the process proceeds to step S9, and n <nsl. If it is, exit the routine.

  In step S9, it is checked whether the yaw rate detection range (maximum value-minimum value) calculated in step S7 is equal to or greater than a threshold value for determining a preset range. This threshold value is used to check whether it is sufficient as a range for obtaining a linear regression equation by least squares, and is set in advance by simulation or the like.

  If the detection range ≧ the threshold value, the process proceeds to step S10. If the detection range <the threshold value, the linear regression equation cannot be obtained accurately, and the routine is exited.

  Thereafter, when proceeding to step S10, based on each function f (γlane, γsensor) on the coordinates of the least square data generated in step S4 described above, a linear regression equation (γsensor) is obtained by the least square method with the estimated yaw rate γlane as a reference. = A · γlane + b).

  Thereafter, the process proceeds to step S11, where the slope a and the intercept b, which are linear regression parameters of the linear regression equation (γsensor = a · γlane + b), are obtained, and the scale factor error and the offset error are calculated based on the slope a and the intercept b. To do. The scale factor error corresponds to the slope a of the linear regression equation (γsensor = a · γlane + b), and the offset error corresponds to the intercept b. Therefore, the offset error is identified as b, while the scale factor error is identified as a. The processes in steps S10 and S11 correspond to the linear regression parameter calculation means and parameter identification means of the present invention.

  Next, the process proceeds to step S12, and it is checked whether the scale factor error a and the offset error b are equal to or less than preset threshold values asl and bsl. The threshold values asl and bsl are limit values that can obtain a correct value by calibrating the sensor yaw rate γsensor, and are determined and set in advance from a simulation or the like. If a ≦ asl and b ≦ bsl, the process proceeds to step S13. If at least one of a> asl and b> bsl, the error is large and it is difficult to obtain a good feedback characteristic even after calibration, so the process jumps to step S14.

Then, when proceeding to step S13, the calibration yaw rate γcal obtained by calibrating the sensor yaw rate γsensor is
γcal ← (1 / a) ・ γsensor-b (9)
To go to step S14. In this case, the scale factor error is defined as (1 / a), the offset error is defined as (−b),
a ′ = (1 / a)
b ′ = (− b)
Then, the equation (9) is
γcal ← a ′ · γsensor + b ′ (9 ′)
It becomes.

  As shown in FIG. 8, when the estimated yaw rate γlane and the sensor yaw rate γsensor have the same value, the linear regression equation (γsensor = a · γlane + b) is a straight line passing through the origin of a = 1 and b = 0 (ideal value). On the other hand, when an error occurs, for example, as shown in the figure as a measured value, the linear regression equation (γsensor = a · γlane + b) becomes a shifted value of a = 0.5 and b = −1. Therefore, in this case, a good feedback characteristic can be obtained by neutrally calibrating the sensor yaw rate γsensor with the equation (9) or (9 ′) to the calibration yaw rate cal passing through the origin.

  Thereafter, when the process proceeds from step S12 or step S13 to step S14, all the least square data (γlane, γsensor) stored in the memory are cleared, and the routine is exited.

  Then, the difference Δγ (Δγ = γ * −γcal) between the target yaw rate γ * obtained by the target yaw rate computing unit 41 and the calibration yaw rate γcal obtained by the calibration yaw rate computing unit 44 is read into the feedback control value computing unit 45. Be turned.

  The feedback control value calculation unit 45 obtains the feedback control value Mzfb by PID control or the like based on the above-described difference Δγ.

  Then, the target yaw moment Mz * is obtained from the difference between the feedforward value Mzff obtained by the feedforward control value computing unit 42 and the feedback control value Mzfb obtained by the feedback control value computing unit 45. The target yaw moment Mz * is read into the rate limiter calculation unit 46.

  The rate limiter calculation unit 46 obtains a difference between the target yaw moment Mz * (n-1) obtained during the previous routine execution and the target yaw moment Mz * (n) obtained this time, and this difference is set in advance. Check if the threshold is exceeded. This threshold value is used to check whether the step between the target yaw moment Mz * (n-1) and the target yaw moment Mz * (n) is large. If the step is large, smooth control cannot be performed. The upper limit value of the target yaw moment Mz * (n) obtained this time is limited.

  Then, this target yaw moment Mz * (n) is read into the steering torque calculator 47. The steering torque calculation unit 47 obtains the steering torque Tp corresponding to the target yaw moment Mz * and outputs it to the motor drive unit 21 to drive the electric motor 12 in a predetermined manner.

  As described above, in this embodiment, the estimated yaw rate γlane is obtained based on the lane line model obtained from the image recognized by the forward recognition device 31, and the estimated yaw rate γlane and the sensor yaw rate γsensor are on the coordinate plane. A function f (γlane, γsensor) is plotted, and a linear regression equation (γsensor = a · γlane + b) is obtained by least squares based on the function f (γlane, γsensor).

  Since the slope a of the linear regression equation is used as the scale factor error and the intercept b is identified as the offset error, the scale factor error a and the offset error b can be detected individually, and neutral calibration can be performed. It can be performed with high accuracy. Furthermore, since the sensor yaw rate γsensor is calibrated with the errors a and b to obtain the calibration yaw rate cal, a highly accurate feedback characteristic can be obtained.

  Further, since the estimated yaw rate γlane is obtained based on the image photographed by the forward recognition device 31, it is not affected by not only the straight road traveling but also the crosswind gradient traveling and the crosswind during traveling. The sensor yaw rate γsensor can be neutrally calibrated with high accuracy.

  The present invention is not limited to the above-described embodiment. For example, in the failure diagnosis apparatus, when one of the scale factor error a and the offset error b exceeds the above-described threshold values asl and bsl. Failure diagnosis means for determining a failure of the yaw rate sensor 34 may be provided.

31 ... Forward recognition device,
32 ... Vehicle speed sensor,
33 ... Steering angle sensor,
34 ... Yaw rate sensor,
41 ... Target yaw rate calculation unit,
43 ... Estimated yaw rate calculation unit,
44. Calibration yaw rate calculation unit,
a: Scale factor error,
b: Offset error,
γcal… calibrated yaw rate,
dγlane / dt ... estimated yaw angular acceleration,
M ... own vehicle,
n: Number of sample data,
V ... Vehicle speed,
γ ... Yaw rate,
γ * ... Target yaw rate,
Mz *: Target yaw moment,
γcal… calibrated yaw rate,
γlane ... Estimated lane yaw rate,
γsensor… sensor yaw rate,
θyaw ... Yaw angle to lane

Claims (6)

Physical quantity measuring means for measuring the first physical quantity;
A physical quantity estimating means different from the physical quantity measuring means for setting a second physical quantity for estimating the first physical quantity;
Coordinate data acquisition means for acquiring coordinate data using the first physical quantity and the second physical quantity as coordinate axes;
Parameter computing means for obtaining a linear regression parameter by a least square method based on the acquired coordinate data;
Parameter identifying means for identifying the slope and intercept of the linear regression parameter as a scale factor error and an offset error;
A measurement apparatus comprising: physical quantity calibration means for calibrating the first physical quantity measured by the physical quantity measurement means based on the scale factor error and the offset error to obtain a calibration physical quantity. The measurement apparatus according to claim 1, wherein the physical quantity calibration unit obtains the calibration physical quantity by subtracting the offset error from a value obtained by dividing the first physical quantity by the scale factor error. The measurement apparatus according to claim 1, wherein the physical quantity calibration unit obtains the calibration physical quantity by adding the offset error to a value obtained by multiplying the first physical quantity by the scale factor error. Lane marking detection means for detecting a lane marking that defines the traveling lane;
A lane yaw angle calculation means for obtaining a curvature and an angle formed by the vehicle and the lane line based on the lane line detected by the lane line detection unit;
The first physical quantity measuring means is a yaw rate sensor for detecting a yaw rate of the vehicle;
An estimated yaw rate calculation unit that estimates a vehicle yaw rate from a lane curvature and a lane yaw angle detected by the lane marking detection unit by the second physical quantity measurement unit;
The coordinate data acquisition unit acquires the coordinate data using the yaw rate detected by the yaw rate sensor and the estimated yaw rate calculated by the estimated yaw rate calculation unit as coordinate axes. The measuring device according to item 1. 5. The measuring apparatus according to claim 4, wherein the scale factor error and the offset error are linear regression parameters for calibrating the yaw rate detected by the yaw rate sensor. The scale factor error and the offset error are compared with a preset failure determination threshold, and when at least one of the errors exceeds the failure determination threshold, the failure is determined as a failure of the physical quantity measuring unit A failure diagnosis apparatus using the measurement apparatus according to claim 1, further comprising a diagnosis unit.

Images