Vehicle road adhesion coefficient estimation method and device

Abstract

The invention provides a vehicle road adhesion coefficient estimation method and device, and the method comprises the steps: setting related parameters of a vehicle, and outputting the rotation speed of wheels of the vehicle; building a three-degree-of-freedom vehicle model, inputting the wheel rotating speed into the three-degree-of-freedom vehicle model, and obtaining different vehicle state quantities under a three-degree-of-freedom vehicle dynamic model; a Dugoff tire model is constructed, the vehicle state quantity is input into the Dugoff tire model, the longitudinal force and the lateral force of the vehicle are obtained, and a road adhesion coefficient expression is established; and adopting an MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient. According to the method, the unscented Kalman filter algorithm is improved by adopting the maximum correlation entropy criterion, the three-degree-of-freedom vehicle dynamics model fused with the Dugoff tire model is established, the road adhesion coefficient is estimated, and the robustness and the estimation precision of the estimated value of the road adhesion coefficient are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of road adhesion coefficient estimation, and in particular to a vehicle road adhesion coefficient estimation method and device. Background Art

[0002] Estimation of the key parameter of road adhesion coefficient in vehicle active safety system. Traditional Kalman filter has poor robustness and accuracy in estimating vehicle road adhesion coefficient in non-Gaussian environment. Summary of the invention

[0003] The present invention aims to provide a method and device for estimating a vehicle-road adhesion coefficient that overcomes the above-mentioned problems or at least partially solves the above-mentioned problems.

[0004] To achieve the above object, the technical solution of the present invention is specifically implemented as follows:

[0005] One aspect of the present invention provides a method for estimating a vehicle-road adhesion coefficient, comprising:

[0006] Set the relevant parameters of the vehicle and output the wheel speed of the vehicle;

[0007] Building a three-degree-of-freedom vehicle model, inputting the wheel speed into the three-degree-of-freedom vehicle model, and obtaining different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model;

[0008] Constructing a Dugoff tire model, inputting the vehicle state quantity into the Dugoff tire model, obtaining the longitudinal force and the lateral force of the vehicle, and establishing an expression for the road adhesion coefficient;

[0009] The MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient.

[0010] Optionally, building a three-degree-of-freedom vehicle model includes:

[0011] Construct the three-degree-of-freedom vehicle dynamics model equations:

[0012]

[0013]

[0014] Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track;x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

[0015] Optionally, constructing the Dugoff tire model includes:

[0016] Construct the Dugoff tire model equation:

[0017]

[0018] Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, C y is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

[0019] Optionally, the using the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient includes:

[0020] A maximum correlation entropy unscented Kalman filter is established, and a modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion, and the vehicle state quantity is updated and corrected to obtain an estimated value of the road adhesion coefficient.

[0021] Optionally, the establishing of the maximum correlation entropy unscented Kalman filter, obtaining a modified covariance matrix based on the maximum correlation entropy criterion, updating and correcting the vehicle state quantity, and obtaining an estimated value of the road adhesion coefficient includes:

[0022] The state equation and measurement equation of the nonlinear system are established by using the vehicle state quantity, the longitudinal force of the vehicle and the lateral force of the vehicle:

[0023] Rewrite the state equation and the measurement equation to obtain a rewritten state space equation and a rewritten measurement equation:

[0024]

[0025]

[0026] Among them, the state variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; wk is the system noise; v k To measure noise; Indicates the longitudinal force of the left front tire, Left front tire lateral force, represents the longitudinal force of the right front tire, The lateral force of the right front tire, represents the longitudinal force of the left rear tire, The lateral force of the left rear tire, represents the longitudinal force of the right rear tire, The matrix corresponding to the lateral force of the right rear tire and the road adhesion coefficient is:

[0027]

[0028] Where:

[0029] Optionally, the using the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient includes:

[0030] 2n+1 Sigma sampling points are obtained through UT transformation;

[0031] Calculate the one-step prediction of 2n+1 Sigma point sets;

[0032] Calculate the transformed state prediction value and covariance matrix through the state equation;

[0033] Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix;

[0034] The modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion;

[0035] The mean and covariance of the system prediction are obtained by weighted summation;

[0036] Calculate Kalman gain;

[0037] Compute the state update and covariance update of the system.

[0038] Optionally, the modified covariance matrix obtained based on the maximum correlation entropy criterion includes:

[0039] Re-set the nonlinear regression model;

[0040] Under the MCC criterion, the modified covariance is obtained by maximizing the cost function;

[0041] The corrected measurement covariance is calculated.

[0042] Another aspect of the present invention provides a vehicle-road adhesion coefficient estimation device, comprising:

[0043] A setting module is used to set relevant parameters of the vehicle and output the wheel speed of the vehicle;

[0044] A vehicle model module is used to build a three-degree-of-freedom vehicle model, input the wheel speed into the three-degree-of-freedom vehicle model, and obtain different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model;

[0045] A tire model module is used to construct a Dugoff tire model, input the vehicle state quantity into the Dugoff tire model, obtain the longitudinal force and the lateral force of the vehicle, and establish an expression for the road adhesion coefficient;

[0046] The estimation module is used to obtain an estimated value of the road adhesion coefficient by using the MCUKF estimation algorithm.

[0047] Optionally, the vehicle model module builds a three-degree-of-freedom vehicle model in the following manner:

[0048] Construct the three-degree-of-freedom vehicle dynamics model equations:

[0049]

[0050] Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track; x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

[0051] Optionally, the tire model module constructs the Dugoff tire model in the following manner:

[0052] Construct the Dugoff tire model equation:

[0053]

[0054] Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, Cy is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

[0055] Optionally, the estimation module uses the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient in the following manner:

[0056] A maximum correlation entropy unscented Kalman filter is established, and a modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion, and the vehicle state quantity is updated and corrected to obtain an estimated value of the road adhesion coefficient.

[0057] Optionally, the estimation module establishes a maximum correlation entropy unscented Kalman filter in the following manner, obtains a modified covariance matrix based on a maximum correlation entropy criterion, updates and corrects the vehicle state quantity, and obtains an estimated value of the road adhesion coefficient:

[0058] The state equation and measurement equation of the nonlinear system are established by using the vehicle state quantity, the longitudinal force of the vehicle and the lateral force of the vehicle:

[0059] Rewrite the state equation and the measurement equation to obtain a rewritten state space equation and a rewritten measurement equation:

[0060]

[0061] Among them, the state variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; w k is the system noise; v k To measure noise; Indicates the longitudinal force of the left front tire, Left front tire lateral force, represents the longitudinal force of the right front tire, The lateral force of the right front tire, represents the longitudinal force of the left rear tire, The lateral force of the left rear tire, represents the longitudinal force of the right rear tire, The matrix corresponding to the lateral force of the right rear tire and the road adhesion coefficient is:

[0062]

[0063] Where:

[0064] Optionally, the estimation module uses the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient in the following manner:

[0065] 2n+1 Sigma sampling points are obtained through UT transformation;

[0066] Calculate the one-step prediction of 2n+1 Sigma point sets;

[0067] Calculate the transformed state prediction value and covariance matrix through the state equation;

[0068] Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix;

[0069] The modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion;

[0070] The mean and covariance of the system prediction are obtained by weighted summation;

[0071] Calculate Kalman gain;

[0072] Compute the state update and covariance update of the system.

[0073] Optionally, the estimation module obtains the modified covariance matrix based on the maximum relevant entropy criterion in the following manner:

[0074] Re-set the nonlinear regression model;

[0075] Under the MCC criterion, the modified covariance is obtained by maximizing the cost function;

[0076] The corrected measurement covariance is calculated.

[0077] It can be seen that, through the vehicle road adhesion coefficient estimation method and device provided by the present invention, the maximum correlation entropy criterion (MCC) is adopted to improve the unscented Kalman (UKF) filter algorithm (MCUKF), a three-degree-of-freedom vehicle dynamics model integrating the Dugoff tire model is established, and the road adhesion coefficient is estimated, thereby improving the robustness and estimation accuracy of the estimated value of the road adhesion coefficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0079] Figure 1 A flow chart of a method for estimating a vehicle-road adhesion coefficient provided by an embodiment of the present invention;

[0080] Figure 2 A flowchart of MCUKF calculation is provided for an embodiment of the present invention;

[0081] Figure 3 A schematic diagram of simulation results provided by an embodiment of the present invention;

[0082] Figure 4 A schematic diagram of the structure of a vehicle-road adhesion coefficient estimation device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0083] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0084] The core of the invention is to set the vehicle parameters and the road adhesion coefficient in the vehicle dynamics software Carsim, output the wheel speed, and then connect the established three-degree-of-freedom vehicle dynamics model in Simulink, combined with the Dugoff tire model, derive the state space expression of the road adhesion coefficient, and then use the MCUKF estimation algorithm to estimate the road adhesion coefficient.

[0085] Figure 1 FIG. 1 is a flow chart showing a method for estimating a vehicle-road adhesion coefficient according to an embodiment of the present invention. Figure 1 The vehicle-road adhesion coefficient estimation method provided by the embodiment of the present invention includes:

[0086] S1, set the relevant parameters of the vehicle and output the wheel speed of the vehicle.

[0087] Specifically, the present invention can output the wheel speed of the vehicle by setting vehicle-related parameters and road adhesion coefficient in Carsim. The vehicle-related parameters are shown in Table 1.

[0088] Table 1 Carsim vehicle parameters

[0089]

[0090]

[0091] S2, building a three-degree-of-freedom vehicle model, inputting the wheel speed into the three-degree-of-freedom vehicle model, and obtaining different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model.

[0092] Specifically, the present invention is based on the kinematic characteristics of the vehicle and takes into account that the state estimation data of the vehicle requires three degrees of freedom, namely the dynamic equations of the longitudinal, lateral and yaw aspects to participate in the calculation. Therefore, the present invention can build a vehicle model in the matlab / simulink simulation software.

[0093] As an optional implementation of the embodiment of the present invention, building a three-degree-of-freedom vehicle model includes:

[0094] Construct the three-degree-of-freedom vehicle dynamics model equations:

[0095]

[0096] Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track; x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

[0097] When specifically implemented, the three-degree-of-freedom vehicle dynamics model equation of the present invention is as follows:

[0098]

[0099] Where: a x 、a y , r are the longitudinal acceleration, lateral acceleration, and yaw rate of the vehicle’s center of mass, respectively; m is the vehicle’s mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a and b are the distances from the center of mass to the front axle and the center of mass to the rear axle, respectively; t f ,t r F is the front axle track and rear axle track respectively; x 、F y They are the longitudinal force and lateral force of the tires respectively. 1, 2, 3, and 4 are the left front wheel, right front wheel, left rear wheel, and right rear wheel respectively.

[0100] S3, constructing a Dugoff tire model, inputting the vehicle state quantity into the Dugoff tire model, obtaining the longitudinal force and the lateral force of the vehicle, and establishing an expression for the road adhesion coefficient.

[0101] Specifically, the data points obtained in step S2 are used as input to the Dugoff tire model to obtain the longitudinal force and lateral force of the vehicle, thereby obtaining expressions for the longitudinal force, lateral force and road adhesion coefficient, providing a theoretical basis for using the maximum correlation entropy unscented Kalman filter.

[0102] As an optional implementation of the embodiment of the present invention, the constructing of the Dugoff tire model includes:

[0103] Construct the Dugoff tire model equation:

[0104]

[0105] Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, C y is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

[0106] In specific implementation, the tire longitudinal force and lateral force of a single wheel can be expressed as:

[0107]

[0108] Where μ is the road adhesion coefficient; F z is the vertical load of the wheel; C x , C y are the longitudinal stiffness and lateral stiffness of the tire respectively. Their values ​​under different vertical loads can be obtained by looking up the table; s is the tire slip (rotation) rate; α is the tire slip angle.

[0109] The critical value L is used to describe the nonlinear characteristics caused by tire slip:

[0110]

[0111]

[0112] Where ε is the speed influence factor, which is used to correct the influence of tire slip speed on tire force; v x is the longitudinal velocity of the wheel.

[0113] In order to ignore the influence of the road adhesion coefficient and facilitate the determination of the coefficient matrix in the algorithm, the Dugoff normalized tire model is introduced:

[0114]

[0115] In the formula, are the longitudinal and lateral normalized tire forces, respectively.

[0116] S4, using the MCUKF estimation algorithm, obtains the estimated value of the road adhesion coefficient.

[0117] As an optional implementation of an embodiment of the present invention, the use of the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient includes: establishing a maximum correlation entropy unscented Kalman filter, obtaining a corrected covariance matrix based on the maximum correlation entropy criterion, updating and correcting the vehicle state quantity, and obtaining an estimated value of the road adhesion coefficient.

[0118] Specifically, the present invention uses the data points estimated in step S2 and the tire force obtained in step S3 to establish a maximum correlation entropy unscented Kalman filter, obtains a corrected covariance matrix based on the maximum correlation entropy criterion, performs self-update and correction of the state quantity, and finally obtains an estimated value of the road adhesion coefficient.

[0119] The step of establishing a maximum correlation entropy unscented Kalman filter, obtaining a modified covariance matrix based on the maximum correlation entropy criterion, updating and correcting the vehicle state quantity, and obtaining an estimated value of the road adhesion coefficient includes:

[0120] The state equation and measurement equation of the nonlinear system are established by using the vehicle state quantity, the longitudinal force of the vehicle and the lateral force of the vehicle:

[0121] Rewrite the state equation and the measurement equation to obtain a rewritten state space equation and a rewritten measurement equation:

[0122]

[0123]

[0124] Among them, the state variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; w k is the system noise; v k To measure noise; Indicates the longitudinal force of the left front tire, Left front tire lateral force, represents the longitudinal force of the right front tire, The lateral force of the right front tire, represents the longitudinal force of the left rear tire, Left rear tire lateral force, represents the longitudinal force of the right rear tire, The matrix corresponding to the lateral force of the right rear tire and the road adhesion coefficient is:

[0125]

[0126] Where:

[0127] In the specific implementation, the maximum correlation entropy unscented Kalman filter is established, and its state variables, control variables and observation variables are as follows:

[0128] The state equation and measurement equation of the general form of the Kalman filter are as follows:

[0129]

[0130] z k =h(x k ,u k +v k ) (9)

[0131] In the formula, x k is the state variable, u k is the control variable; Z k is the measurement output; w k is the system noise; v k To measure noise.

[0132] Rewrite the vehicle's state space equation and measurement equation as follows:

[0133]

[0134] In the formula,

[0135]

[0136] State variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; w k is the system noise; v k To measure noise.

[0137] The matrix corresponding to the road adhesion coefficient of the observer is as follows:

[0138]

[0139] The initial input value of the maximum correlation entropy unscented Kalman filter is established as follows:

[0140] The initial value of the system's state variables X0 = [0.90.90.90.9]T; the process noise covariance matrix Q = 0.1diag [0.1040.1040.1040.104]; the measurement noise covariance matrix R = 0.01diag [0.2050.2050.205].

[0141] As an optional implementation of the embodiment of the present invention, the MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient, which includes:

[0142] 2n+1 Sigma sampling points are obtained through UT transformation;

[0143] Calculate the one-step prediction of 2n+1 Sigma point sets;

[0144] Calculate the transformed state prediction value and covariance matrix through the state equation;

[0145] Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix;

[0146] The modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion;

[0147] The mean and covariance of the system prediction are obtained by weighted summation;

[0148] Calculate Kalman gain;

[0149] Compute the state update and covariance update of the system.

[0150] Among them, the modified covariance matrix based on the maximum correlation entropy criterion includes:

[0151] Re-set the nonlinear regression model;

[0152] Under the MCC criterion, the modified covariance is obtained by maximizing the cost function;

[0153] The corrected measurement covariance is calculated.

[0154] Specifically, the MCUKF calculation process includes:

[0155] Step 1: Get 2n+1 Sigma sampling points through UT transformation:

[0156]

[0157] In the formula, Represents the matrix (n+λ)P k-1|k-1 The i-th column of the square root, n represents the dimension of the state vector, and λ represents the scaling parameter.

[0158] Step 2: Calculate the one-step prediction of 2n+1 Sigma point sets:

[0159]

[0160] Step 3: Calculate the transformed state prediction value through the state equation and the covariance matrix P k|k-1 :

[0161]

[0162] Step 4: Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix:

[0163]

[0164] Calculate the predicted observations through the observation equation:

[0165]

[0166] Step 5: Based on the maximum correlation entropy criterion, the corrected covariance matrix is ​​as follows:

[0167] The nonlinear regression model can be reformulated as:

[0168]

[0169] in, T k By k The Cholesky decomposition of is obtained.

[0170] Under the MCC criterion, the modified covariance is obtained by maximizing the cost function

[0171]

[0172] Get the corrected measurement covariance:

[0173]

[0174] Step 6: Obtain the mean and covariance of the system prediction through weighted summation:

[0175]

[0176]

[0177] Step 7: Calculate the Kalman gain:

[0178]

[0179] Step 8: Calculate the state update and covariance update of the system:

[0180]

[0181] The MCUKF algorithm flow chart is as follows: Figure 2 As shown, the process noise w k and measurement noise v k satisfy the non-Gaussian noise distribution, and

[0182] In a specific implementation, the MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient, which includes:

[0183] S1, initialize the state variable X 0|0 , covariance matrix P 0|0 , process noise covariance matrix Q and measurement noise covariance matrix R;

[0184] S2, UT transformation is performed by the following formula to obtain 2n+1 Sigma sampling points;

[0185]

[0186] S3, calculate the weights of 2n+1 Sigma point sets using the following formula:

[0187]

[0188] S4, calculate the predicted value of the state variable by the following formula and covariance P k|k-1 , use Cholesky decomposition to get Tk|k-1,P:

[0189]

[0190] S5, calculate the new 2n+1 Sigma sampling points by the following formula to obtain the new sampling point set and predicted observation value

[0191]

[0192] S6, the corrected measurement covariance is obtained based on the maximum correlation entropy criterion through the following formula

[0193]

[0194] S7, calculate the Kalman gain matrix K by the following formula k :

[0195]

[0196] The updated system state variables are calculated by the following formula and the covariance matrix P k|k :

[0197]

[0198]

[0199] S8, determine whether k>N, if yes, output the estimated value of the road adhesion coefficient, if not, set k=k+1 and return to execute step S4.

[0200] As an optional implementation of the embodiment of the present invention, a result simulation comparison is performed.

[0201] In Carsim, the road adhesion coefficient is set to 0.4, and the simulation results are as follows: Figure 3 shown.

[0202] It can be seen that, through the vehicle road adhesion coefficient estimation method provided by the present invention, the maximum correlation entropy criterion (MCC) is adopted to improve the unscented Kalman (UKF) filter algorithm (MCUKF), a three-degree-of-freedom vehicle dynamics model integrating the Dugoff tire model is established, and the road adhesion coefficient is estimated, thereby improving the robustness and estimation accuracy of the estimated value of the road adhesion coefficient.

[0203] Figure 4 The schematic diagram of the structure of the vehicle-road adhesion coefficient estimation device provided by the embodiment of the present invention is shown. The vehicle-road adhesion coefficient estimation device applies the above method. The following is only a brief description of the structure of the vehicle-road adhesion coefficient estimation device. For other matters not covered, please refer to the relevant description in the above vehicle-road adhesion coefficient estimation method. Figure 4 , the vehicle-road adhesion coefficient estimation device provided by the embodiment of the present invention comprises:

[0204] A setting module is used to set relevant parameters of the vehicle and output the wheel speed of the vehicle;

[0205] A vehicle model module is used to build a three-degree-of-freedom vehicle model, input the wheel speed into the three-degree-of-freedom vehicle model, and obtain different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model;

[0206] A tire model module is used to construct a Dugoff tire model, input the vehicle state quantity into the Dugoff tire model, obtain the longitudinal force and the lateral force of the vehicle, and establish an expression for the road adhesion coefficient;

[0207] The estimation module is used to obtain an estimated value of the road adhesion coefficient using the MCUKF estimation algorithm.

[0208] As an optional implementation of the embodiment of the present invention, the vehicle model module builds a three-degree-of-freedom vehicle model in the following manner:

[0209] Construct the three-degree-of-freedom vehicle dynamics model equations:

[0210]

[0211] Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track; x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

[0212] As an optional implementation of the embodiment of the present invention, the tire model module constructs the Dugoff tire model in the following manner:

[0213] Construct the Dugoff tire model equation:

[0214]

[0215] Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, C y is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

[0216] As an optional implementation of the embodiment of the present invention, the estimation module uses the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient in the following manner:

[0217] A maximum correlation entropy unscented Kalman filter is established, and a modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion, and the vehicle state quantity is updated and corrected to obtain an estimated value of the road adhesion coefficient.

[0218] As an optional implementation of the embodiment of the present invention, the estimation module establishes a maximum correlation entropy unscented Kalman filter in the following manner, obtains a modified covariance matrix based on the maximum correlation entropy criterion, updates and corrects the vehicle state quantity, and obtains an estimated value of the road adhesion coefficient:

[0219] The state equation and measurement equation of the nonlinear system are established by using the vehicle state quantity, the longitudinal force of the vehicle and the lateral force of the vehicle:

[0220] Rewrite the state equation and the measurement equation to obtain a rewritten state space equation and a rewritten measurement equation:

[0221]

[0222] Among them, the state variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; w k is the system noise; v k To measure noise; Indicates the longitudinal force of the left front tire, Left front tire lateral force, represents the longitudinal force of the right front tire, The lateral force of the right front tire, represents the longitudinal force of the left rear tire, Left rear tire lateral force, represents the longitudinal force of the right rear tire, The matrix corresponding to the lateral force of the right rear tire and the road adhesion coefficient is:

[0223]

[0224] Where:

[0225] As an optional implementation of the embodiment of the present invention, the estimation module uses the MCUKF estimation algorithm to obtain an estimated value of the road adhesion coefficient in the following manner:

[0226] 2n+1 Sigma sampling points are obtained through UT transformation;

[0227] Calculate the one-step prediction of 2n+1 Sigma point sets;

[0228] Calculate the transformed state prediction value and covariance matrix through the state equation;

[0229] Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix;

[0230] The modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion;

[0231] The mean and covariance of the system prediction are obtained by weighted summation;

[0232] Calculate Kalman gain;

[0233] Compute the state update and covariance update of the system.

[0234] As an optional implementation of the embodiment of the present invention, the estimation module obtains the modified covariance matrix based on the maximum correlation entropy criterion in the following manner:

[0235] Re-set the nonlinear regression model;

[0236] Under the MCC criterion, the modified covariance is obtained by maximizing the cost function;

[0237] The corrected measurement covariance is calculated.

[0238] It can be seen that, through the vehicle road adhesion coefficient estimation device provided by the present invention, the maximum correlation entropy criterion (MCC) is adopted to improve the unscented Kalman (UKF) filter algorithm (MCUKF), a three-degree-of-freedom vehicle dynamics model integrating the Dugoff tire model is established, and the road adhesion coefficient is estimated, thereby improving the robustness and estimation accuracy of the estimated value of the road adhesion coefficient.

[0239] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for estimating a vehicle-road adhesion coefficient, characterized in that: include: Set the relevant parameters of the vehicle and output the wheel speed of the vehicle; Building a three-degree-of-freedom vehicle model, inputting the wheel speed into the three-degree-of-freedom vehicle model, and obtaining different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model; Constructing a Dugoff tire model, inputting the vehicle state quantity into the Dugoff tire model, obtaining the longitudinal force and the lateral force of the vehicle, and establishing an expression for the road adhesion coefficient; The MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient.

2. The method according to claim 1, characterized in that The three-degree-of-freedom vehicle model is constructed as follows: Construct the three-degree-of-freedom vehicle dynamics model equation: Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track; x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

3. The method according to claim 1, characterized in that The construction of the Dugoff tire model comprises: Construct the Dugoff tire model equation: Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, C y is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

4. The method according to claim 3, characterized in that The MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient, including: A maximum correlation entropy unscented Kalman filter is established, and a modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion, and the vehicle state quantity is updated and corrected to obtain an estimated value of the road adhesion coefficient.

5. The method according to claim 4, characterized in that The method of establishing a maximum correlation entropy unscented Kalman filter, obtaining a modified covariance matrix based on a maximum correlation entropy criterion, updating and correcting the vehicle state quantity, and obtaining an estimated value of the road adhesion coefficient includes: The state equation and measurement equation of the nonlinear system are established by using the vehicle state quantity, the longitudinal force of the vehicle and the lateral force of the vehicle: Rewrite the state equation and the measurement equation to obtain a rewritten state space equation and an improved measurement equation: Among them, the state variable x k =[μ1μ2μ3μ4] T , control variables Z k is the measurement output; w k is the system noise; v k To measure noise; Indicates the longitudinal force of the left front tire, Left front tire lateral force, represents the longitudinal force of the right front tire, The lateral force of the right front tire, represents the longitudinal force of the left rear tire, The lateral force of the left rear tire, represents the longitudinal force of the right rear tire, The matrix corresponding to the lateral force of the right rear tire and the road adhesion coefficient is: Where:

6. The method according to claim 4, characterized in that The MCUKF estimation algorithm is used to obtain the estimated value of the road adhesion coefficient, including: 2n+1 Sigma sampling points are obtained through UT transformation; Calculate the one-step prediction of 2n+1 Sigma point sets; Calculate the transformed state prediction value and covariance matrix through the state equation; Calculate a new set of 2n+1 Sigma points by predicting the state mean and covariance matrix; The modified covariance matrix is ​​obtained based on the maximum correlation entropy criterion; The mean and covariance of the system prediction are obtained by weighted summation; Calculate Kalman gain; Compute the state update and covariance update of the system.

7. The method according to claim 6, characterized in that The modified covariance matrix based on the maximum correlation entropy criterion includes: Re-set the nonlinear regression model; Under the MCC criterion, the modified covariance is obtained by maximizing the cost function; The corrected measurement covariance is calculated.

8. A vehicle-road adhesion coefficient estimation device, characterized in that: include: A setting module is used to set relevant parameters of the vehicle and output the wheel speed of the vehicle; A vehicle model module is used to build a three-degree-of-freedom vehicle model, input the wheel speed into the three-degree-of-freedom vehicle model, and obtain different vehicle state quantities under the three-degree-of-freedom vehicle dynamics model; A tire model module is used to construct a Dugoff tire model, input the vehicle state quantity into the Dugoff tire model, obtain the longitudinal force and the lateral force of the vehicle, and establish an expression for the road adhesion coefficient; The estimation module is used to obtain an estimated value of the road adhesion coefficient using the MCUKF estimation algorithm.

9. The device according to claim 8, characterized in that The vehicle model module builds a three-degree-of-freedom vehicle model in the following way: Construct the three-degree-of-freedom vehicle dynamics model equation: Among them, a x is the longitudinal acceleration of the vehicle center of mass, a y is the lateral acceleration of the vehicle's center of mass, r is the vehicle's yaw rate; m is the vehicle's mass; I z is the moment of inertia of the vehicle around the z-axis; δ is the front wheel turning angle; a is the distance from the center of mass to the front axle, b is the distance from the center of mass to the rear axle; t f is the front axle track, t r F is the rear axle track; x is the tire longitudinal force, F y is the lateral force of the tire, 1 is the left front wheel, 2 is the right front wheel, 3 is the left rear wheel, and 4 is the right rear wheel.

10. The device according to claim 8, characterized in that The tire model module constructs the Dugoff tire model in the following way: Construct the Dugoff tire model equation: Among them, F x is the longitudinal force of a single tire, F y is the lateral force of a single tire; μ is the road adhesion coefficient; is the normalized longitudinal tire force, is the normalized lateral tire force; F z is the vertical load of the wheel; C x is the longitudinal stiffness of the tire, C y is the tire cornering stiffness; s is the tire slip (turn) rate; α is the tire slip angle, and L is the critical value, which is used to describe the nonlinear characteristics caused by tire slip.

Images