Vehicle state amount estimation device, and vehicle control device

Abstract

According to one aspect of a vehicle state amount estimation device and a vehicle control device according to the present invention, a sensor coordinate system acceleration and a sensor coordinate system angular velocity output by an inertial sensor are subjected to coordinate transformation to obtain a horizontal coordinate system acceleration and a horizontal coordinate system angular velocity, an estimated speed of a vehicle is obtained by integrating an arithmetic expression of a horizontal coordinate system including the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity, and a vehicle body slip angle is estimated on the basis of the estimated speed. Thus, a reduction in accuracy of the estimation of the vehicle body slip angle based on the output of the inertial sensor can be suppressed even when the inertial sensor is mounted in a tilted orientation on the vehicle.

Description

Vehicle state quantity estimation device and vehicle control device

[0001] The present invention relates to a vehicle state quantity estimating device and a vehicle control device.

[0002] The vehicle state quantity estimation device of Patent Document 1 includes an inertial sensor that detects acceleration, angular velocity, etc., a state quantity estimation unit that estimates the state quantity of the vehicle based on the detection value of the inertial sensor and preset parameters such as mass and center of gravity position, and an external environment recognition means that detects the relative displacement and angle between the vehicle and structures around the vehicle, and updates the parameters during a period when the detection value of the external environment recognition means is stable.

[0003] Japanese Patent Application Laid-Open No. 2018-024265

[0004] However, when the vehicle body slip angle is estimated based on an estimated speed obtained by integrating the acceleration and angular velocity detected by an inertial sensor mounted on the vehicle, if the inertial sensor is mounted at an angle to the vehicle, an integration error occurs in the estimated speed, resulting in a problem of reduced accuracy in estimating the vehicle body slip angle.

[0005] Therefore, an object of the present invention is to provide a vehicle state quantity estimation device and a vehicle control device that can prevent a decrease in the estimation accuracy of the vehicle body slip angle based on the output of an inertial sensor even when the inertial sensor is attached to the vehicle at an angle.

[0006] Therefore, in one aspect, the vehicle state quantity estimating device according to the present invention includes a vehicle state estimating unit that acquires a sensor coordinate system acceleration and a sensor coordinate system angular velocity of the vehicle that are defined in a sensor coordinate system output by an inertial sensor mounted on the vehicle, performs coordinate transformation on the sensor coordinate system acceleration and the sensor coordinate system angular velocity, determines a horizontal coordinate system acceleration and a horizontal coordinate system angular velocity of the vehicle that are defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration, and calculates an estimated speed of the vehicle by integrating an arithmetic expression for the horizontal coordinate system that includes the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity, and a vehicle body slip angle estimating unit that estimates a vehicle body slip angle of the vehicle based on the estimated speed.

[0007] In one aspect, the vehicle control device according to the present invention acquires a sensor coordinate system acceleration and a sensor coordinate system angular velocity of the vehicle, which are defined in a sensor coordinate system output by an inertial sensor mounted on the vehicle, performs coordinate transformation on the sensor coordinate system acceleration and the sensor coordinate system angular velocity, determines a horizontal coordinate system acceleration and a horizontal coordinate system angular velocity of the vehicle, which are defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration, determines an estimated speed of the vehicle by integrating an arithmetic expression for the horizontal coordinate system including the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity, estimates a vehicle body slip angle of the vehicle based on the estimated speed, and executes motion control of the vehicle using the vehicle body slip angle.

[0008] According to the present invention, even when the inertial sensor is mounted on the vehicle at an angle, it is possible to prevent a decrease in the accuracy of estimating the vehicle body slip angle based on the output of the inertial sensor.

[0009] 1 is a block diagram showing the functions of a vehicle control device including a vehicle state quantity estimating device. FIG. 2 is a block diagram showing input / output signals in a vehicle state estimating section and a vehicle body slip angle estimating section. FIG. 3 is a flowchart showing the flow of calculation for estimating a vehicle body slip angle. FIG. 4 is a block diagram showing the functions of an observation quantity vector calculating section. FIG. 5 is a block diagram showing the functions of a vehicle speed flag calculating section. FIG. 6 is a block diagram showing the functions of an initial roll angle / initial pitch angle calculating section. FIG. 7 is a flowchart showing the flow of calculation processing for initial roll angle / initial pitch angle. FIG. 8 is a block diagram showing the functions of a vehicle body slip angle calculating section. FIG. 9 is a block diagram showing the functions of a vehicle state estimating section. FIG. 10 is a diagram showing the relationship between mathematical expressions used to estimate a vehicle body slip angle. FIG. 11 is a block diagram showing the functions of an estimating section. FIG. 12 is a block diagram showing the functions of a correcting section. FIG. 13 is a block diagram showing the functions of a Kalman gain calculating section. FIG. 14 is a block diagram showing the functions of an "R" selecting section. FIG. 15 is a block diagram showing the functions of a switching section. FIG. 16 is a block diagram showing the functions of a corrected estimated state quantity vector switching section. FIG. 17 is a block diagram showing the functions of a corrected estimated covariance matrix switching section. FIG. 18 is a block diagram showing the functions of a delay / reset section. FIG. 19 is a block diagram showing the functions of a corrected estimated state quantity delay / reset section. FIG. 19 is a block diagram showing the functions of a corrected estimated covariance delay / reset section. FIG. 19 is a flowchart showing the flow of calculation processing for a corrected estimated state quantity vector and a corrected estimated covariance matrix. FIG. 19 is a flowchart showing the process of correcting an estimated longitudinal speed. 10 is a flowchart showing a process for correcting an estimated lateral speed.

[0010] 1 is a block diagram showing the functions of a vehicle control device 20 including a vehicle state quantity estimating device 90. In the vehicle state quantity estimating device 90, a vehicle state quantity estimating device 90 is provided.

[0011] The vehicle 1 is equipped with a steering actuator 11 that applies a steering force to the front wheels, which are the steered wheels of the vehicle 1, in response to the driver's steering operation, an inertia sensor 12 that detects the acceleration and angular velocity of the vehicle 1, a wheel speed sensor 13 that detects the rotational speed of each of the four wheels of the vehicle 1, i.e., the left and right front wheels and the left and right rear wheels, and an ESC actuator 14 that realizes ESC (Electronic Stability Control) that distributes braking force to the four wheels to stabilize the vehicle behavior in the event of skidding of the vehicle 1.

[0012] The inertial sensor 12 is a six-axis inertial sensor that detects accelerations along three axes of the vehicle 1, namely, longitudinal, lateral, and vertical, as well as angular velocities along the three axes, i.e., longitudinal acceleration, lateral acceleration, vertical acceleration, roll rate, pitch rate, and yaw rate. The vehicle control device 20 mounted on the vehicle 1 has a microcomputer 20A as a control unit that outputs the results of calculations based on input information.

[0013] The microcomputer 20A includes a nonvolatile memory, and runs a program stored in the nonvolatile memory to estimate the vehicle body slip angle of the vehicle 1. The microcomputer 20A then uses the estimated vehicle body slip angle to perform motion control of the vehicle 1, such as anti-skid control. In other words, the microcomputer 20A functions as a vehicle state quantity estimating device 90 that estimates the vehicle body slip angle of the vehicle 1.

[0014] 1, the observation quantity vector calculation unit 30, the initial roll angle / initial pitch angle calculation unit 40, the vehicle speed flag calculation unit 50, the vehicle state estimation unit 60, and the vehicle body slip angle estimation unit 70 are functional units that constitute the vehicle state quantity estimation device 90. Also, the ESC logic unit 80 shown in FIG. 1 is a functional unit that executes motion control of the vehicle 1 (more specifically, skid prevention logic) based on the estimated value of the vehicle body slip angle.

[0015] The vehicle state estimation unit 60 acquires the signals of the sensor coordinate system acceleration and the sensor coordinate system angular velocity defined in the sensor coordinate system output by the inertial sensor 12, and calculates the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration by coordinate transformation. Furthermore, the vehicle state estimation unit 60 calculates the estimated speed of the vehicle 1 by integrating an arithmetic expression for the horizontal coordinate system including the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity.

[0016] The vehicle body slip angle estimator 70 then estimates the vehicle body slip angle of the vehicle 1 based on the estimated speed determined by the vehicle state estimator 60. The ESC logic unit 80 acquires a signal of the vehicle body slip angle estimated by the vehicle body slip angle estimator 70 and outputs a signal of required braking force to the ESC actuator 14, thereby executing motion control that stabilizes the behavior of the vehicle 1. As will be described in detail later, the coordinate transformation performed by the vehicle state estimator 60 is a process for preventing integral errors from occurring in the estimated speed even if the inertial sensor 12 is mounted at an angle on the vehicle 1, and can improve the estimation accuracy of the vehicle body slip angle.

[0017] The following describes each functional unit that constitutes the vehicle state quantity estimation device 90. The observable vector calculation unit 30 calculates the average wheel speed of the driven wheels of the vehicle 1. For example, if the vehicle 1 is a front-wheel drive vehicle, the observable vector calculation unit 30 calculates the average rear wheel speed, which is the average value of the wheel speeds of the left and right rear wheels.

[0018] Furthermore, the observation quantity vector calculation unit 30 calculates an observation quantity vector (in other words, an observed longitudinal speed) from the wheel speeds of the driven wheels, and sets the observation quantity vector (in other words, an observed lateral speed) to 0. Then, the observation quantity vector calculation unit 30 outputs signals of the rear wheel average wheel speed and the observation quantity vector to the vehicle state estimating unit 60, and outputs a signal of the rear wheel average wheel speed to the vehicle speed flag calculation unit 50.

[0019] The vehicle speed flag calculation unit 50 acquires the signal of the rear wheel average wheel speed calculated by the observation quantity vector calculation unit 30, and determines whether a vehicle speed flag indicating that the vehicle 1 is moving, more specifically, whether the vehicle 1 is in a moving state or a stopped state, is on or off based on the rear wheel average wheel speed. The vehicle speed flag is set to on when the vehicle 1 is moving, and is set to off when the vehicle 1 is in a stopped state. The vehicle speed flag calculation unit 50 then outputs the on / off signal of the vehicle speed flag to the vehicle state estimation unit 60 and the vehicle body slip angle estimation unit 70.

[0020] Here, the average wheel speed used by vehicle speed flag calculation unit 50 to calculate the on / off state of the vehicle speed flag is the average wheel speed of the rear wheels, which are driven wheels, and the difference between the vehicle speed (in other words, vehicle body speed) and the wheel speed of the driven wheels is small. Therefore, vehicle speed flag calculation unit 50 can accurately determine whether the vehicle speed flag is on / off, i.e., whether vehicle speed is occurring, and can improve the accuracy of the process of estimating vehicle body slip angle using the vehicle speed flag.

[0021] When the vehicle speed flag calculated by the vehicle speed flag calculation unit 50 is in the OFF state (i.e., the vehicle 1 is in the stopped state), the initial roll angle / initial pitch angle calculation unit 40 calculates an estimated roll angle and an estimated pitch angle based on the longitudinal acceleration, lateral acceleration, and vertical acceleration included in the sensor coordinate system acceleration output from the inertial sensor 12. Then, the initial roll angle / initial pitch angle calculation unit 40 holds the estimated roll angle and estimated pitch angle when the vehicle speed flag changes from the OFF state to the ON state as initial values ​​(more specifically, the initial roll angle and the initial pitch angle) to be used in the coordinate transformation, and outputs the initial values ​​to the vehicle state estimation unit 60.

[0022] 2 is a block diagram showing input / output signals to and from the vehicle state estimator 60 and the vehicle body slip angle estimator 70. The vehicle state estimator 60 receives the acceleration signal and angular velocity signal (in other words, input quantities) output from the inertial sensor 12, the observation quantity vector (more specifically, the observed longitudinal velocity and observed lateral velocity), the initial roll angle / initial pitch angle, the average rear wheel speed, and the vehicle speed flag signals, and calculates a post-switching modified estimated state quantity vector (more specifically, the estimated longitudinal velocity in the horizontal coordinate system and the estimated lateral velocity in the horizontal coordinate system) based on these signals. The vehicle body slip angle estimator 70 then receives the post-switching modified estimated state quantity vector signal and the vehicle speed flag signal from the vehicle state estimator 60, calculates an estimated vehicle body slip angle based on these signals, and outputs a signal of the calculated estimated vehicle body slip angle.

[0023] 3 is a flowchart showing the flow of calculations for estimating vehicle state quantities and vehicle body slip angle, executed by the microcomputer 20A. In step S101, the microcomputer 20A acquires the outputs of the inertia sensor 12 and the wheel speed sensor 13. Next, in step S102 (observation quantity vector calculation unit 30), the microcomputer 20A calculates the average rear wheel speed (in other words, the average driven wheel speed) from the output of the wheel speed sensor 13.

[0024] In step S103 (vehicle speed flag calculation unit 50), microcomputer 20A calculates whether a vehicle speed flag is on or off. In step S104 (initial roll angle / initial pitch angle calculation unit 40), microcomputer 20A calculates and stores an initial roll angle and an initial pitch angle, which are initial values ​​of the estimated roll angle and estimated pitch angle.

[0025] Then, in step S105 (vehicle state estimating section 60), the microcomputer 20A calculates a post-switching processing corrected estimated state quantity vector (more specifically, the estimated longitudinal velocity and the estimated lateral velocity in the horizontal coordinate system) based on the output of the inertial sensor 12, the observed longitudinal velocity and the observed lateral velocity, which are observation quantity vectors, the initial roll angle / initial pitch angle, the rear wheel average speed, and the vehicle speed flag signals. Next, in step S106 (vehicle body slip angle estimating section 70), the microcomputer 20A calculates the vehicle body slip angle based on the post-switching processing corrected estimated state quantity vector (more specifically, the estimated longitudinal velocity and the estimated lateral velocity in the horizontal coordinate system).

[0026] Each of the above-mentioned functional units will be described in more detail below. Fig. 4 is a block diagram showing the functions of the observation quantity vector calculation unit 30. The observation quantity vector calculation unit 30 acquires signals of the rotational speeds (wheel speeds) of the wheels of the vehicle 1 from the wheel speed sensors 13.

[0027] The averaging processing unit 31 included in the observable vector calculation unit 30 performs processing to calculate the average value of the rotational speeds of the driven wheels, i.e., the driven wheel average wheel speed, among the rotational speeds of the wheels of the vehicle 1. In one aspect of the present embodiment, the vehicle 1 is a front-wheel drive four-wheel vehicle, and the averaging processing unit 31 calculates the rear wheel average wheel speed, which is the average value of the rotational speeds of the left and right rear wheels. The observable vector calculation unit 30 also outputs the observable vector based on the rear wheel average wheel speed.

[0028] 5 is a block diagram showing the functions of the vehicle speed flag calculation unit 50. The vehicle speed flag calculation unit 50 has a comparison unit 51. The comparison unit 51 compares the rear wheel average speed obtained by the observation quantity vector calculation unit 30 with a threshold value V , which is set to a very small speed in order to distinguish between the stopped state and the running state of the vehicle 1. Limit Compare with.

[0029] The comparison unit 51 then compares the rear wheel average speed with the threshold value V Limit On the other hand, when the rear wheel average speed is higher than the threshold value V Limit Hereinafter, when it is determined that the vehicle 1 is in a stopped state, the vehicle speed flag is turned off.

[0030] 6 is a block diagram showing the functions of the initial roll angle / initial pitch angle calculation unit 40. The initial roll angle / initial pitch angle calculation unit 40 has functional units, namely, a low-pass filter 41, an initial value calculation unit 42, and a signal holding unit 43. The low-pass filter 41 acquires the signals of the longitudinal acceleration ax, the lateral acceleration ay, and the vertical acceleration az from the inertial sensor 12, and attenuates frequency components higher than a cutoff frequency contained in the signals of the longitudinal acceleration ax, the lateral acceleration ay, and the vertical acceleration az.

[0031] Then, the initial value calculation unit 42 calculates an initial roll angle φ according to Equation 1 based on the longitudinal acceleration ax, lateral acceleration ay, and vertical acceleration az that have passed through the low-pass filter 41. ini (hat) and initial pitch angle θ ini (hat) is calculated.

[0032] The signal holding unit 43 receives the initial roll angle φ calculated by the initial value calculation unit 42.ini (hat) and initial pitch angle θ ini When the vehicle speed flag is on, indicating that the vehicle 1 is in a traveling state, the signal holding unit 43 holds the initial roll angle φ ini (hat) and initial pitch angle θ ini (Hat) signal is held.

[0033] On the other hand, when the vehicle speed flag is off, indicating that the vehicle 1 is in a stopped state, the signal holding unit 43 holds the initial roll angle φ ini (hat) and initial pitch angle θ ini (hat) signal is not held, but is successively updated to the calculation result of the initial value calculation unit 42. In other words, when the vehicle 1 is in a stopped state, the signal holding unit 43 outputs the initial roll angle φ ini (hat) and initial pitch angle θ ini The signal (hat) is successively updated to the calculation result in the initial value calculation unit 42, and when the vehicle 1 starts traveling, the initial roll angle φ ini (hat) and initial pitch angle θ ini In this manner, when the vehicle speed flag is in the OFF state, that is, when the vehicle 1 is stopped, the initial roll angle / initial pitch angle calculation unit 40 calculates the initial roll angle and initial pitch angle from the relational expressions between the accelerations in the sensor coordinate system and the accelerations in the horizontal coordinate system.

[0034] 7 is a flowchart showing the flow of calculation processing in the initial roll angle / initial pitch angle calculation unit 40. In step S111, the microcomputer 20A calculates the initial roll angle φ ini (hat) and initial pitch angle θ ini (hat) is calculated.

[0035] Next, in step S112, the microcomputer 20A acquires the vehicle speed flag output by the vehicle speed flag calculation unit 50. Then, in step S113, the microcomputer 20A determines whether the vehicle speed flag is on or off.

[0036] Here, if the vehicle speed flag is on and the running state of the vehicle 1 is detected, the microcomputer 20A returns to step S111 to set the initial roll angle φ ini (hat) and initial pitch angle θ ini On the other hand, if the vehicle speed flag is off and the vehicle 1 is detected to be in a stopped state, the microcomputer 20A proceeds to step S114, and updates the initial roll angle φ calculated in step S111 this time (in other words, calculated immediately before the vehicle transitions from a traveling state to a stopped state). ini (hat) and initial pitch angle θ ini Hold (hat).

[0037] 8 is a block diagram showing the functions of the vehicle body slip angle estimator 70. The vehicle body slip angle estimator 70 has an angle calculator 71 and an output switcher 72. The angle calculator 71 receives the corrected estimated state quantity vector (estimated longitudinal velocity u in the horizontal coordinate system) after switching processing from the vehicle state estimator 60. H (hat) and estimated lateral velocity v H (hat)) signal is acquired. In this specification, the subscript "H" indicates a value in a horizontal coordinate system perpendicular to the direction of gravitational acceleration, as will be described in detail later.

[0038] Then, the angle calculation unit 71 calculates the estimated vehicle body slip angle β (hat) in accordance with Equation 2. The corrected estimated state quantity vector after the switching process, which the angle calculation unit 71 acquires from the vehicle state estimation unit 60, is switched depending on whether the vehicle speed flag is on or off, as will be described in detail later, and uses an estimated value that is updated sequentially when the vehicle 1 is traveling, and uses an appropriate initial state quantity when the vehicle 1 is stopped.

[0039] Depending on whether the vehicle speed flag acquired from the vehicle speed flag calculation unit 50 is on or off, the output switching unit 72 switches the output value of the estimated vehicle body slip angle β (hat) between the value calculated by the angle calculation unit 71 and 0. Here, when the vehicle speed flag is on and a traveling state of the vehicle 1 is detected, the output switching unit 72 outputs the estimated vehicle body slip angle β (hat) calculated by the angle calculation unit 71, and when the vehicle speed flag is off and a stopped state of the vehicle 1 is detected, the output switching unit 72 outputs the estimated vehicle body slip angle β (hat) as 0.

[0040] The ESC logic unit 80 executes motion control of the vehicle 1 based on the signal of the estimated vehicle body slip angle β (hat) acquired from the output switching unit 72. As shown in Fig. 1 , the ESC logic unit 80 has a reference vehicle model 81, a comparison unit 82, and a gain unit 83. The reference vehicle model 81 is constructed as, for example, a two-wheel vehicle model.

[0041] The reference vehicle model 81 acquires signals of the steering angle of the steered wheels and the average rear wheel speed (in other words, the vehicle speed), and outputs signals of a reference yaw rate and a reference vehicle body slip angle that are estimated to occur in the vehicle 1 under the conditions of the steering angle and the average rear wheel speed. The comparison unit 82 compares the reference yaw rate output by the reference vehicle model 81 with the actual yaw rate detected by the inertial sensor 12, and determines the control deviation of the yaw rate.

[0042] Furthermore, a comparison unit 82 compares the reference vehicle body slip angle output by the reference vehicle model 81 with the estimated vehicle body slip angle β (hat) calculated by the vehicle body slip angle estimation unit 70 to calculate the control deviation of the vehicle body slip angle. A gain unit 83 generates a required brake pressure signal for each wheel based on the control deviation of the yaw rate and the vehicle body slip angle, and outputs the required brake pressure signal to the ESC actuator 14 to stabilize the vehicle behavior, that is, to distribute braking force so as to bring the actual yaw rate and actual vehicle body slip angle closer to the reference yaw rate and reference vehicle body slip angle.

[0043] FIG. 9 is a block diagram showing the functions of the vehicle state estimating unit 60. FIG. 10 is a diagram showing the relationship between mathematical expressions used to estimate the vehicle body slip angle. The vehicle state estimating unit 60 acquires the output signal (input quantity) of the inertial sensor 12, the observation quantity vector, the initial roll angle / initial pitch angle, the average rear wheel speed, and the vehicle speed flag, and calculates the corrected estimated state quantity vector after the switching process (corrected estimated longitudinal speed u in the horizontal coordinate system). H (hat) and the corrected estimated lateral velocity v H (hat)) to the vehicle body slip angle estimator 70. The vehicle state estimator 60 has the following functional parts: an estimator 61, a corrector 62, a switcher 63, and a delay / reset unit 64.

[0044] 11 is a block diagram showing the functions of the estimator 61. The estimator 61 has a state quantity estimator 61A and a covariance estimator 61B. The state quantity estimator 61A acquires the detection output of the inertial sensor 12 and the previous corrected estimated state quantity vector, and determines and outputs the current corrected estimated state quantity vector. The covariance estimator 61B acquires the detection output of the inertial sensor 12, the previous corrected estimated state quantity vector, and the previous estimated covariance matrix, and determines and outputs an estimated covariance matrix that indicates the relationship between the detection output of the inertial sensor 12 and the corrected estimated state quantity vector.

[0045] The estimation unit 61 estimates the state quantity xk (hat) according to Equation 3. The estimation unit 61 estimates the covariance matrix Pk according to Equation 4. In Equation 4, Ak is the system matrix and Q is the covariance matrix of the process noise.

[0046] If the mounting position of the inertial sensor 12 relative to the vehicle 1 is tilted, it will be affected by the component force components of the gravitational acceleration, and integral errors will occur in the estimated values ​​of the longitudinal velocity and lateral velocity based on the detection output (sensor coordinate system values) of the inertial sensor 12. Therefore, the estimation unit 61 converts the signals of the sensor coordinate system acceleration and sensor coordinate system angular velocity defined in the sensor coordinate system output by the inertial sensor 12 into horizontal coordinate system acceleration and horizontal coordinate system angular velocity defined in a horizontal coordinate system perpendicular to the direction of the gravitational acceleration, and uses these horizontal coordinate system acceleration and horizontal coordinate system angular velocity to find an estimated state quantity vector and an estimated covariance matrix.

[0047] That is, the equation of motion in the sensor coordinate system is expressed as Equation 5. Therefore, if the inertial sensor 12 is mounted at an angle to the vehicle 1 and the longitudinal acceleration ax and lateral acceleration ay are affected by the component force components of gravitational acceleration, this will cause an integration error, resulting in an error in the estimated vehicle body slip angle β (hat) calculated using the longitudinal velocity u (hat) and lateral velocity v (hat) estimated from the detection output of the inertial sensor 12.

[0048] Therefore, the estimation unit 61 converts the signals of the sensor coordinate system acceleration and the sensor coordinate system angular velocity into horizontal coordinate system acceleration and horizontal coordinate system angular velocity defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration, and calculates estimated values ​​using the horizontal coordinate system acceleration and horizontal coordinate system angular velocity as shown in Equation 6, thereby suppressing the occurrence of integration errors.

[0049] The angle calculation unit 71 of the vehicle body slip angle estimator 70 calculates the estimated vehicle body slip angle β (hat) using the acceleration in the horizontal coordinate system, as shown in the above-mentioned Equation 2. Furthermore, the estimation unit 61 estimates the covariance matrix Pk using the horizontal coordinate system acceleration and horizontal coordinate system angular velocity.

[0050] The coordinate transformation process will be described in detail below. When the sensor coordinate system accelerations are defined as longitudinal acceleration ax, lateral acceleration ay, and vertical acceleration az, and the sensor coordinate system angular velocities are defined as roll rate p, pitch rate q, and yaw rate r, the estimation unit 61 performs coordinate transformation shown in Equation 8 using the transformation matrix shown in Equation 7. Then, through the above coordinate transformation, the estimation unit 61 obtains estimated longitudinal acceleration axH (hat), estimated lateral acceleration ayH (hat), and estimated vertical acceleration azH (hat), which are horizontal coordinate system accelerations, as well as roll rate pH (hat), pitch rate qH (hat), and yaw rate rH (hat), which are horizontal coordinate system angular velocities.

[0051]

[0052] Here, in the transformation matrix of Equation 7, φ (hat) is an estimated roll angle, and θ (hat) is an estimated pitch angle, and the estimated roll angle φ (hat) and estimated pitch angle θ (hat) can be calculated using the Euler angle formula as shown in Equation 9.

[0053] Furthermore, the estimation unit 61 uses a rigid body motion model equation that does not include vehicle specifications as an estimation model in the observer logic, thereby making it possible to estimate the vehicle body slip angle with high accuracy regardless of the degree to which the specifications have been identified. The above estimation model (nonlinear model equation) is expressed, for example, by Equation 10.

[0054] Then, the estimation unit 61 sets a transformation matrix shown in Equation 7 based on the output of the estimation model, determines the system matrix A as shown in Equation 11, and estimates the covariance matrix P in accordance with Equation 4 described above using the system matrix A and the previous value P of the covariance matrix P.

[0055] 12 is a block diagram showing the functions of the correction unit 62. The correction unit 62 acquires signals of the output of the inertial sensor 12, the estimated state quantity vector, the estimated covariance matrix, the observation quantity vector, and the previous corrected estimated state quantity vector, and outputs signals of the corrected estimated state quantity vector and the corrected estimated covariance matrix.

[0056] The correction unit 62 has the following functional units: a Kalman gain calculation unit 62A, a state quantity correction unit 62B, and a covariance correction unit 62C. The Kalman gain calculation unit 62A calculates the Kalman gain Kk according to Equation 12. In Equation 12, H is the observation matrix, and R is the covariance matrix of the sensor noise.

[0057] The observation matrix H is determined according to Equation 13.

[0058] The state quantity correcting section 62B corrects the state quantity xk (hat) estimated by the state quantity estimating section 61A of the estimating section 61 in accordance with Equation 14, using the Kalman gain Kk. In addition, in Equation 14, z is an observation quantity vector expressed by Equation 15.

[0059] Furthermore, the covariance correction unit 62C corrects the covariance matrix Pk estimated by the covariance estimation unit 61B of the estimation unit 61 in accordance with Equation 16, using the Kalman gain Kk. As described above, the vehicle body slip angle can be determined with high accuracy even at high slip angles (in other words, in the nonlinear region) by the nonlinear Kalman filter logic that utilizes the nonlinear estimation model formula.

[0060] 13 is a block diagram showing the function of a Kalman gain calculation unit 62A of the correction unit 62. The Kalman gain calculation unit 62A calculates the Kalman gain Kk according to the above-mentioned Equation 12, and has an “R” selection unit 62A1 that selects the covariance matrix R of the sensor noise used in the calculation of the Kalman gain Kk.

[0061] 14 is a block diagram showing the function of the "R" selection unit 62A1 of the Kalman gain calculation unit 62A. The "R" selection unit 62A1 is a functional unit that determines whether or not to correct and obtain the estimated speed depending on the magnitude of the acceleration in the sensor coordinate system. The "R" selection unit 62A1 acquires the longitudinal acceleration ax and the lateral acceleration ay detected by the inertial sensor 12, and also calculates the estimated lateral speed v in the horizontal coordinate system from the previously corrected state quantity vector. H Get (Hat).

[0062] Then, the “R” selection unit 62A1 calculates the absolute value of the longitudinal acceleration ax, and determines whether the absolute value of the longitudinal acceleration ax is equal to or exceeds the threshold ax Limit Here, the “R” selection unit 62A1 determines whether the absolute value of the longitudinal acceleration ax is smaller than the threshold ax Limit If the absolute value of the longitudinal acceleration ax is smaller than the threshold ax, the longitudinal speed element in the covariance matrix R of the sensor noise is set to a first predetermined value (in other words, a small value). Limit If the value is equal to or greater than this, the front / rear speed element in the covariance matrix R of the sensor noise is set to a second predetermined value (in other words, a large value) that is greater than the first predetermined value.

[0063] Further, the "R" selection unit 62A1 calculates the absolute value of the lateral acceleration ay, and the absolute value of the lateral acceleration ay is set to the threshold ay Limit Furthermore, the "R" selection unit 62A1 determines whether the previously corrected estimated lateral velocity v H The absolute value of (hat) is calculated, and the previous corrected estimated lateral velocity v H The absolute value of (hat) is the threshold v HLimit Determine whether it is smaller than

[0064] Then, the "R" selection unit 62A1 selects whether the absolute value of the lateral acceleration ay is equal to the threshold ay Limit and the first condition that the previous corrected estimated lateral velocity v H The absolute value of (hat) is the threshold v HLimit If at least one of the second conditions is satisfied, the lateral velocity element in the covariance matrix R of the sensor noise is set to a first predetermined value (small value). On the other hand, the "R" selection unit 62A1 selects the lateral velocity element in the covariance matrix R of the sensor noise when the absolute value of the lateral acceleration ay is smaller than the threshold ay. Limit Above, and the previous corrected estimated lateral speed v H The absolute value of (hat) is the threshold v HLimit If so, the lateral velocity element in the covariance matrix R of the sensor noise is set to a second predetermined value (large value).

[0065] Here, the "R" selection unit 62A1 calculates a diagonal matrix based on the longitudinal speed elements and lateral speed elements, and outputs the sensor noise covariance matrix R. By repeating the integration processes of the estimation unit 61 using the above model and the correction unit 62 using the sensor and reference values, it is possible to acquire an estimated state quantity vector with high accuracy.

[0066] 15 is a block diagram showing the function of the switching unit 63 (see FIG. 9 ) in the vehicle state estimating unit 60. The switching unit 63 is a functional unit that acquires a corrected estimated state quantity vector, a corrected estimated covariance matrix, an initial roll angle / initial pitch angle, an average rear wheel speed, and a vehicle speed flag, and outputs a corrected estimated state quantity vector after switching processing and a corrected estimated covariance matrix after switching processing.

[0067] The switching unit 63 is a functional unit that distinguishes between a running state and a stopped state using a vehicle speed flag, and uses estimated state quantities and estimated covariance matrices in the running state and appropriate initial state quantities as the state quantities and covariance matrix in the stopped state in order to reduce estimation errors and prevent divergence of estimated values ​​when the vehicle 1 is stopped. The switching unit 63 has a corrected estimated state quantity vector switching unit 63A that switches the state quantity vector, and a corrected estimated covariance matrix switching unit 63B that switches the covariance matrix.

[0068] 16 is a block diagram showing the functions of the corrected estimated state quantity vector switching unit 63A. The corrected estimated state quantity vector switching unit 63A acquires the signal of the corrected estimated state quantity vector output by the modifying unit 62, and also acquires the signal of the vehicle speed flag output by the vehicle speed flag calculating unit 50. Then, when the vehicle 1 is in a traveling state where the vehicle speed flag is on, the corrected estimated state quantity vector switching unit 63A outputs the signal of the corrected estimated state quantity vector acquired from the modifying unit 62.

[0069] Furthermore, when the vehicle 1 is in a stopped state where the vehicle speed flag is off, the corrected estimated state quantity vector switching unit 63A outputs, as a signal of the corrected estimated state quantity vector, a predetermined initial state quantity in place of the output of the corrector 62. The corrected estimated state quantity vector switching unit 63A can output, as the initial state quantities, the corrected estimated longitudinal speed as a value based on the rear wheel average wheel speed, and the corrected estimated lateral speed as zero, and further includes signals of the initial pitch angle and the initial roll angle as the initial state quantities.

[0070] 17 is a block diagram showing the function of the modified estimated covariance matrix switching unit 63B that switches the covariance matrix. The modified estimated covariance matrix switching unit 63B receives the modified estimated covariance matrix signal output by the modifying unit 62 and the vehicle speed flag signal output by the vehicle speed flag calculation unit 50.

[0071] When the vehicle 1 is in a traveling state where the vehicle speed flag is on, the corrected estimated covariance matrix switching unit 63B outputs a signal of the corrected estimated covariance matrix acquired from the corrector 62. When the vehicle 1 is in a stopped state where the vehicle speed flag is off, the corrected estimated covariance matrix switching unit 63B outputs a predetermined fixed value, such as the covariance matrix Q of the process noise, instead of the output of the corrector 62 as the signal of the corrected estimated covariance matrix. The switching unit 63 described above makes it possible to cancel the estimation process at low speeds where state quantity estimation becomes difficult.

[0072] Fig. 18 is a block diagram showing the function of the delay / reset unit 64 shown in Fig. 9. The delay / reset unit 64 is a functional unit that acquires the corrected estimated state quantity vector and corrected estimated covariance matrix after the switching process from the switching unit 63, as well as acquires the initial roll angle / initial pitch angle and vehicle speed flag, and delays and resets the corrected estimated state quantity vector and corrected estimated covariance matrix.

[0073] The delay / reset unit 64 reduces the integration error by resetting the estimated values ​​(corrected estimated state quantity vector, corrected estimated covariance matrix) to predetermined initial values ​​when the vehicle 1 starts moving. The delay / reset unit 64 has a corrected estimated state quantity delay / reset unit 64A that performs delay / reset processing on the corrected estimated state quantity vector, and a corrected estimated covariance delay / reset unit 64B that performs delay / reset processing on the corrected estimated covariance matrix.

[0074] 19 is a block diagram showing the function of the corrected estimated state quantity delay / reset unit 64A. The corrected estimated state quantity delay / reset unit 64A acquires the signal of the corrected estimated state quantity vector after the switching process output by the switching unit 63, the signals of the initial roll angle / initial pitch angle output by the initial roll angle / initial pitch angle calculation unit 40, and the signal of the vehicle speed flag output by the vehicle speed flag calculation unit 50. The corrected estimated state quantity delay / reset unit 64A has a delay element 64A1 that delays the signal of the corrected estimated state quantity vector after the switching process by one sample.

[0075] The delay element 64A1 resets the corrected estimated state quantity vector to a predetermined initial value when the vehicle 1 starts, which is the timing when the vehicle flag is raised. Here, the delay element 64A1 resets (in other words, initializes) the corrected estimated state quantity vector (corrected estimated longitudinal speed, corrected estimated lateral speed) after the switching process to zero when the vehicle 1 starts, and outputs it as a corrected estimated state quantity vector including an initial pitch angle and an initial roll angle.

[0076] 20 is a block diagram showing the function of the corrected estimated covariance delay / reset unit 64B. The corrected estimated covariance delay / reset unit 64B acquires the signal of the corrected estimated covariance matrix after switching processing output by the switching unit 63 and the signal of the vehicle speed flag output by the vehicle speed flag calculation unit 50. The corrected estimated covariance delay / reset unit 64B has a delay element 64B1 that delays the signal of the corrected estimated covariance matrix by one sample.

[0077] When the vehicle 1 starts moving, which is the timing when the vehicle flag is raised, the delay element 64B1 resets the corrected estimated covariance matrix to a predetermined initial value, for example, the covariance matrix Q of the process noise. As described above, the delay / reset unit 64 resets the error accumulation due to the integration process of the nonlinear Kalman filter and feeds it back as the previous value.

[0078] 21 is a flowchart showing the flow of the process of estimating the vehicle state quantities executed by the microcomputer 20A, that is, the process of calculating the corrected estimated state quantity vector and the corrected estimated covariance matrix in the vehicle state estimation unit 60. In step S121, the microcomputer 20A executes the calculation of the estimated state quantity vector and the estimated covariance matrix in the estimation unit 61.

[0079] Next, in step S122, the microcomputer 20A performs calculations of the corrected estimated state quantity vector and the corrected estimated covariance matrix in the correcting unit 62. Next, the microcomputer 20A proceeds to step S123, where it determines whether the vehicle speed flag is on or off.

[0080] When the vehicle 1 is in a traveling state with the vehicle speed flag on, the microcomputer 20A proceeds to step S124 and outputs, as values ​​after the switching process, the corrected estimated state quantity vector and the corrected estimated covariance matrix calculated by the correction unit 62. On the other hand, when the vehicle 1 is in a stopped state with the vehicle speed flag off, the microcomputer 20A proceeds to step S125 and outputs, as values ​​after the switching process, the initial state quantities including the initial pitch angle / initial roll angle, instead of the corrected estimated state quantity vector and the corrected estimated covariance matrix calculated by the correction unit 62.

[0081] Then, in step S126, the microcomputer 20A outputs the corrected estimated state quantity vector after the switching process as the estimated state quantity. The processes of steps S123 to S126 are performed by the switching unit 63.

[0082] Furthermore, in step S127, the microcomputer 20A determines whether the vehicle speed flag has been switched from off to on, that is, whether it is the time of starting the vehicle 1. Then, if the vehicle speed flag has been switched from off to on, the microcomputer 20A proceeds to step S128.

[0083] In step S128, the microcomputer 20A resets (in other words, initializes) the corrected estimated state quantity vector and the corrected estimated covariance matrix after the switching process, which are the outputs of the switching unit 63, and outputs them as the previous corrected estimated state quantity vector and the previous corrected estimated covariance matrix. On the other hand, if the timing has not been such that the vehicle speed flag has been switched from off to on, in other words, if the vehicle 1 has not started moving, the microcomputer 20A proceeds to step S129.

[0084] In step S129, the microcomputer 20A outputs the corrected estimated state quantity vector and corrected estimated covariance matrix after the switching process, which are the outputs of the switching unit 63, as the previous corrected estimated state quantity vector and the previous corrected estimated covariance matrix. The processes of steps S127 to S129 are performed by the delay / reset unit 64.

[0085] 22 is a flowchart showing in detail the process of correcting the estimated longitudinal speed in step S122 (correction unit 62). In step S141, the microcomputer 20A calculates whether the absolute value of the longitudinal acceleration ax is equal to or exceeds the threshold value ax Limit It is determined whether the absolute value of the longitudinal acceleration ax is smaller than the threshold ax Limit If it is smaller than , the process proceeds to step S142.

[0086] In step S142, the microcomputer 20A selects parameters that closely track the reference (rear wheel average speed) by setting the front / rear speed element in the sensor noise covariance matrix R to a first predetermined value (small value). Next, the microcomputer 20A proceeds to step S143, where it corrects the estimated front / rear speed using the rear wheel average speed as the observation quantity vector.

[0087] On the other hand, the absolute value of the longitudinal acceleration ax is the threshold ax LimitIf so, the microcomputer 20A proceeds to step S144. In step S144, the microcomputer 20A sets the front / rear speed element in the covariance matrix R of the sensor noise to a second predetermined value (in other words, a large value) that is greater than the first predetermined value, thereby selecting a reference, that is, a parameter that does not follow the rear wheel average wheel speed. Next, in step S145, the microcomputer 20A cancels the correction of the estimated front / rear speed.

[0088] 23 is a flowchart showing in detail the process of correcting the estimated lateral velocity in step S122 (correction unit 62). In step S151, the microcomputer 20A calculates whether the absolute value of the lateral acceleration ay is equal to or exceeds the threshold ay. Limit and the first condition that the previous corrected estimated lateral velocity v H The absolute value of (hat) is the threshold v HLimit and the second condition that the difference is smaller than or equal to 1.

[0089] If at least one of the first condition and the second condition is satisfied, the microcomputer 20A proceeds to step S152, where it sets the lateral velocity element in the covariance matrix R of the sensor noise to a first predetermined value (in other words, a small value) to select a reference, that is, a parameter that closely follows the estimated lateral velocity. Next, the microcomputer 20A proceeds to step S153, where it selects the estimated lateral velocity v H (hat) = 0 is used as the observation vector, and the estimated lateral velocity v H (Hat) to fix.

[0090] On the other hand, the microcomputer 20A determines whether the absolute value of the lateral acceleration ay is equal to the threshold ay Limit Above, and the previous corrected estimated lateral speed v H The absolute value of (hat) is the threshold v HLimit If the value is equal to or greater than the reference value, the process proceeds to step S154, where the lateral velocity element in the covariance matrix R of the sensor noise is set to a second predetermined value (large value) to obtain the reference, i.e., the estimated lateral velocity v HNext, in step S155, the microcomputer 20A cancels the correction of the estimated lateral speed.

[0091] As described above, the correction unit 62, which uses the sensor and the reference value, determines certain conditions from the longitudinal acceleration detected by the inertial sensor 12 and corrects the estimated longitudinal speed in the horizontal coordinate system using the wheel speed of the driven wheels (for example, the rear wheels in a front-wheel drive vehicle). The correction unit 62 also corrects the estimated lateral speed in the horizontal coordinate system to 0 [m / s] using the lateral acceleration detected by the inertial sensor 12 and the previously corrected estimated lateral speed.

[0092] The technical ideas described in the above embodiments can be used in appropriate combinations as long as no contradictions arise. Furthermore, although the contents of the present invention have been specifically described with reference to preferred embodiments, it is obvious that a person skilled in the art can adopt various modified embodiments based on the basic technical ideas and teachings of the present invention.

[0093] For example, the vehicle 1 may be a rear-wheel drive vehicle, in which case the observable vector calculation unit 30 calculates the average front wheel speed, which is the average of the wheel speeds of the left and right front wheels, which are driven wheels. Also, the inertial sensor 12 may be a sensor composed of an individual acceleration sensor and a gyro sensor.

[0094] Furthermore, motion control of the vehicle 1 using the vehicle slip angle is not limited to ESC, which stabilizes vehicle behavior by distributing braking force to all four wheels. For example, a vehicle control device can use the vehicle slip angle to predict the passing position of the wheels, determine road surface conditions such as the presence or absence of ruts based on the vehicle slip angle, and control the steering angle of the steered wheels based on the vehicle slip angle.

[0095] 1...vehicle, 12...inertial sensor, 20...vehicle control device, 20A...microcomputer (control unit), 30...observation quantity vector calculation unit, 40...initial roll angle / initial pitch angle calculation unit, 50...vehicle speed flag calculation unit, 60...vehicle state estimation unit, 70...vehicle body slip angle estimation unit, 80...ESC logic unit

Claims

1. A vehicle state quantity estimation device comprising: a vehicle state estimation unit that acquires a sensor coordinate system acceleration and a sensor coordinate system angular velocity of the vehicle, which are defined in a sensor coordinate system output by an inertial sensor mounted on the vehicle; coordinate-transforms the sensor coordinate system acceleration and the sensor coordinate system angular velocity to determine a horizontal coordinate system acceleration and a horizontal coordinate system angular velocity of the vehicle, which are defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration; and calculates an estimated speed of the vehicle by integrating an arithmetic expression for the horizontal coordinate system including the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity; and a vehicle body slip angle estimation unit that estimates a vehicle body slip angle of the vehicle based on the estimated speed.

2. A vehicle state quantity estimation device according to claim 1, further comprising: a vehicle speed flag calculation unit that determines whether a vehicle speed flag indicating that the vehicle has moved is on or off; and an initial roll angle / initial pitch angle calculation unit that, when the vehicle speed flag is off, calculates an estimated roll angle and an estimated pitch angle based on the longitudinal acceleration, lateral acceleration, and up / down acceleration included in the sensor coordinate system acceleration output from the inertial sensor, holds the estimated roll angle and estimated pitch angle when the vehicle speed flag changes from off to on as initial values ​​to be used in the coordinate transformation, and outputs the initial values ​​to the vehicle state estimation unit.

3. A vehicle state quantity estimation device according to claim 2, wherein the vehicle speed flag calculation unit determines whether the vehicle speed flag is on or off based on an average wheel speed of the driven wheels of the vehicle obtained from a wheel speed sensor mounted on the vehicle.

4. A vehicle state quantity estimating device according to claim 3, wherein the vehicle state estimating section comprises: an estimating section that estimates an estimated state quantity vector and an estimated covariance matrix based on an input quantity obtained from the inertial sensor, a previously corrected estimated state quantity vector, and a previously corrected estimated covariance matrix; a correcting section that determines a corrected estimated state quantity vector and a corrected estimated covariance matrix based on the input quantity obtained from the inertial sensor, the estimated state quantity vector, the estimated covariance matrix, an observation quantity vector determined based on detection values ​​of the wheel speed sensors, and the previously corrected estimated state quantity vector; and a switching section that determines a corrected estimated state quantity vector and a corrected estimated covariance matrix based on the corrected estimated state quantity vector, the corrected estimated covariance matrix, the initial value, the average wheel speed of the driven wheels, and the vehicle speed flag, and outputs the corrected estimated state quantity vector to the vehicle body slip angle estimating section. a delay / reset unit that calculates the previously corrected estimated state vector and the previously corrected estimated covariance matrix based on the corrected estimated state vector, the corrected estimated covariance matrix, the initial value, and the vehicle speed flag.

5. A vehicle state quantity estimation device according to claim 4, wherein the correction unit determines whether or not to obtain the estimated speed by correction depending on the magnitude of the sensor coordinate system acceleration included in the input quantity obtained from the inertial sensor.

6. A vehicle control device comprising a control unit that outputs a result of calculation based on input information, wherein the control unit: acquires a sensor coordinate system acceleration and a sensor coordinate system angular velocity of the vehicle defined in a sensor coordinate system output by an inertial sensor mounted on the vehicle; performs coordinate transformation on the sensor coordinate system acceleration and the sensor coordinate system angular velocity to determine a horizontal coordinate system acceleration and a horizontal coordinate system angular velocity of the vehicle defined in a horizontal coordinate system perpendicular to the direction of gravitational acceleration; determines an estimated speed of the vehicle by integrating an arithmetic expression for the horizontal coordinate system including the horizontal coordinate system acceleration and the horizontal coordinate system angular velocity; estimates a body slip angle of the vehicle based on the estimated speed; and performs motion control of the vehicle using the body slip angle.

Images