A fault-tolerant control method for distributed steer-by-wire vehicles

Abstract

The present invention proposes a fault-tolerant control method for a distributed steer-by-wire vehicle, comprising: establishing a distributed steer-by-wire vehicle dynamics model, a two-degree-of-freedom reference model, and a steering motor failure model; in the design of an upper-level fault-tolerant controller, an adaptive terminal sliding mode control method is adopted, which has the characteristics of strong robustness and adaptive parameter changes, and can ensure target tracking performance while taking into account the robustness of the system. In the design of a lower-level torque and angle optimization allocation reconstruction controller, a failure factor is introduced, and two more objective functions are established on the basis of minimizing the tire load rate, ensuring that the optimization objective function has a solution while reducing the error between the longitudinal and lateral forces and the desired target values, as well as the error between the yaw moment and the additional yaw moment. The present invention fully considers the uncertainty of the system when the steering fails, and can ensure target tracking performance while taking into account the robustness of the system, so as to improve the handling stability of the distributed steer-by-wire vehicle when the steering fails.

Description

Technical Field

[0001] The present invention relates to the field of automobile assisted driving, and in particular to a fault-tolerant control method for a distributed steer-by-wire vehicle. Background Art

[0002] The distributed steer-by-wire system has two sets of steering actuators on the front and rear axles (front and rear axle type) or four sets of steering actuators on four wheels (four-wheel distributed). Compared with front-wheel steer-by-wire, the distributed steer-by-wire system is free from the mechanical connection constraints between the left and right wheels, and can provide a wider range of lateral control force. The wheel angle is no longer restricted by the trapezoidal mechanical structure, and the steering is more flexible, which improves the vehicle's maneuverability and flexibility. It exhibits more precise, agile and stable steering performance during the steering process, and can complete special motion conditions such as low-speed steering, on-the-spot steering, oblique driving, and translation.

[0003] Distributed steer-by-wire systems have multiple actuators. As the number of actuators increases, the probability of actuator failure also increases. Furthermore, distributed steer-by-wire systems are susceptible to harsh environments, which can lead to structural damage. Therefore, there is an urgent need to introduce fault-tolerant control into distributed steer-by-wire systems to mitigate the impact of actuator failures on the handling stability of distributed steer-by-wire vehicles. Currently, control methods such as conventional sliding mode control, PID control, model predictive control, and neural network control are used in distributed steer-by-wire vehicle handling stability control. While conventional sliding mode control offers advantages such as simplicity and robustness, it suffers from chattering, making it unsuitable for distributed steer-by-wire vehicle stability control. PID control, while simple in design and flexible in structure, is primarily used in single-input, single-output systems. Model predictive control linearizes the vehicle model, reducing the computational complexity of the control process and improving real-time control performance. However, this also reduces control accuracy, and its computational burden increases with the complexity of the system's dynamics. Neural network control can approximate external disturbances and modeling errors, and has strong robustness and fault tolerance, but it has disadvantages such as slow learning speed and long calculation time, and it is difficult to prove the convergence and stability of the learning process in an analytical way.

[0004] Currently, most torque and angle distribution reconstruction controllers of distributed steer-by-wire systems only use minimizing the tire load rate as the optimization objective. Under complex working conditions, no solution may occur, requiring repeated solutions. This leads to problems such as long optimization solution time and slow tire force distribution response. Summary of the Invention

[0005] The purpose of the present invention is to propose a fault-tolerant control method for a distributed steer-by-wire vehicle; it fully considers the uncertainty of the system when the steering fails, and can take into account the robustness of the system while ensuring target tracking performance, so as to improve the handling stability of the distributed steer-by-wire vehicle when the steering fails.

[0006] To achieve the above object, the technical solution of the present invention is: a distributed steer-by-wire vehicle fault-tolerant control method, comprising the following steps:

[0007] S1. Establish a distributed steer-by-wire vehicle dynamics model;

[0008] S2. Establish a two-degree-of-freedom vehicle reference model;

[0009] S3. determining a desired longitudinal speed and an ideal lateral speed;

[0010] S4. Establishing a failure model for the steering motor of a distributed steer-by-wire vehicle;

[0011] S5. Establishing an upper adaptive terminal sliding mode fault-tolerant controller for a distributed steer-by-wire vehicle;

[0012] S6. Establish a distributed steer-by-wire vehicle lower-layer torque and steering angle optimization distribution reconstruction controller.

[0013] Preferably, the establishment of a distributed steer-by-wire vehicle dynamics model is specifically as follows:

[0014] Longitudinal dynamic equation:

[0015]

[0016] Lateral dynamics equation:

[0017]

[0018] Yaw dynamics model:

[0019]

[0020] Among them are:

[0021]

[0022] Where m is the vehicle mass; v x and v y are the components of the vehicle's center of mass velocity on the x-axis and y-axis respectively; a x is the longitudinal acceleration; a y is the lateral acceleration; ω r is the vehicle's yaw rate; β is the vehicle's center of mass sideslip angle; I z is the yaw moment of inertia of the vehicle around the z axis; l f 、l r are the distances from the center of mass of the vehicle to the front and rear axles, respectively; b is half the wheelbase; ij = fl, fr, rl, rr represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively; F tyij 、Ftxij are tire lateral force and tire longitudinal force respectively; F xij 、F yij F txij 、F tyij The longitudinal force and lateral force of each wheel decomposed and synthesized in the vehicle coordinate system; F x 、F y F txij 、F tyij The total longitudinal force and total lateral force decomposed and synthesized in the vehicle coordinate system; M z is the yaw moment of the vehicle; δ ij is the wheel angle;

[0023] Establishing the wheel dynamics model:

[0024]

[0025] Where J is the moment of inertia; ω ij is the angular velocity of each wheel; T ij is the driving torque; F zij is the vertical load; f ij is the rolling resistance coefficient; r ij is the rolling radius;

[0026] According to the moment balance relationship, the vertical load on each wheel is calculated as follows:

[0027]

[0028] Where g is the acceleration due to gravity; h is the height of the center of mass from the ground;

[0029] Building the Magic Formula Tire Model:

[0030] Y(x)=Dsin{Carctan[Bx-E(Bx-arctan(Bx))]} (7)

[0031] Where Y(x) is the lateral force F of the tire tyij Or the longitudinal force F txij ; x is the tire slip angle α ij Or tire slip rate λ ij ; B, C, D, and E are fitting parameters in the tire characteristic curve, which are stiffness factor, curve shape factor, curve peak factor, and curve curvature factor, respectively.

[0032] Preferably, the tire slip rate of each wheel is calculated as:

[0033]

[0034] Where u ij is the wheel center speed of each wheel; ωij is the angular velocity of each wheel; r is the rolling radius of the wheel;

[0035] The wheel center speed of each wheel can be expressed as:

[0036]

[0037] Preferably, the slip angles of the four wheels are calculated as:

[0038]

[0039] Preferably, the establishment of a two-degree-of-freedom vehicle reference model is specifically as follows:

[0040]

[0041] Where k f is the front axle equivalent cornering stiffness of the reference model; k r is the equivalent cornering stiffness of the rear axle of the reference model; δ is the front wheel turning angle of the reference model;

[0042] The ideal yaw rate of a car in steady-state driving is as follows:

[0043]

[0044] Where, ω rd is the ideal yaw rate of the car when it is driving in a steady state, and K is the stability factor:

[0045]

[0046] The expected yaw rate must satisfy the following formula:

[0047]

[0048] Where μ represents the road adhesion coefficient; the ideal yaw rate of the reference model based on equations (12) and (14) can be expressed as:

[0049]

[0050] Preferably, the desired longitudinal velocity is expressed as:

[0051]

[0052] Where, v xd is the desired longitudinal velocity; v x0 is the longitudinal speed at the initial moment; a xd is the target longitudinal acceleration.

[0053] Preferably, the ideal lateral vehicle speed is expressed as:

[0054] vyd =v x tanβ d (17)

[0055] Where, v yd is the ideal lateral speed, β d is the ideal center of mass sideslip angle.

[0056] Preferably, the establishment of a distributed steer-by-wire vehicle steering motor failure model is specifically as follows:

[0057] The output torque expression of the steering motor fault is:

[0058] u λ =ζu d +Δu (18)

[0059] Where ζ∈[0,1] is the steering motor failure factor; u λ is the fault output torque of the steering motor; u d is the expected output torque of the steering motor; Δu is other faults of the steering motor;

[0060] When ζ = 0 and Δu = 0, the steering motor fails completely, the brake mechanism in the steering system takes effect, and the wheel maintains the wheel angle when the steering motor fails completely, and fault-tolerant control is performed by adding yaw torque.

[0061] Preferably, the establishment of the upper layer adaptive terminal sliding mode fault-tolerant controller in S5 includes the following steps:

[0062] S5.1. Modify the two-degree-of-freedom vehicle reference model to a four-wheel independent steering two-degree-of-freedom vehicle model and add an additional yaw moment ΔM:

[0063]

[0064] S5.2. Establish the tracking error of the additional yaw moment adaptive terminal sliding mode controller:

[0065] e r =ω r -ω rd (20)

[0066] S5.3. Establish the sliding mode surface of the additional yaw moment adaptive terminal sliding mode controller:

[0067]

[0068] Where k is an adaptive parameter, c>0, q and p are positive odd numbers, and the expression of k is:

[0069]

[0070] Wherein, σ>0,ξ>0,0<ε<1; w is any real number; e is the natural logarithm;

[0071] S5.4. Use the exponential reaching law to switch the additional yaw moment adaptive terminal sliding mode controller:

[0072]

[0073] Where η>0,κ>0

[0074] S5.5. Combining Equations (19) to (23), we can obtain the final control law of the adaptive terminal sliding mode controller:

[0075]

[0076] Where ΔM is the additional yaw moment, and the positive and negative values ​​represent the vehicle's steering direction, with a positive value indicating a left turn and a negative value indicating a right turn. S5.6. Establish the tracking error of the longitudinal speed adaptive terminal sliding mode controller for the distributed steer-by-wire vehicle:

[0077] e x =v x -v xd (25)

[0078] S5.7. Establish the sliding mode surface of the longitudinal speed adaptive terminal sliding mode controller for the distributed steer-by-wire vehicle:

[0079]

[0080] Where, c x is a controller parameter with a value greater than zero, q x 、p x is a positive odd number, k x The expression is:

[0081]

[0082] Where σ x >0,ξ x >0,0<ε x <1;w x is any real number; e is the natural logarithm;

[0083] S5.8. Use the exponential reaching law to switch the longitudinal vehicle speed adaptive terminal sliding mode controller:

[0084]

[0085] Where η x , κ x are controller parameters with values ​​greater than zero;

[0086] S5.9. Combining equations (1), (25) and (28), we can obtain the final control law of the longitudinal vehicle speed adaptive terminal sliding mode controller:

[0087]

[0088] Where, F xd is the expected total longitudinal force;

[0089] S5.10. Establish a distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller with a tracking error of:

[0090] e y =v y -v yd (30)

[0091] S5.11. Establish the sliding mode surface of the distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller:

[0092]

[0093] Where, c y is a controller parameter with a value greater than zero, q y 、p y is a positive odd number, k y The expression is:

[0094]

[0095] Where σ y >0,ξ y >0,0<ε y <1;w y is any real number; e is the natural logarithm;

[0096] S5.12. Use the exponential reaching law to switch the lateral speed adaptive terminal sliding mode controller:

[0097]

[0098] Where η y , κ y are controller parameters with values ​​greater than zero;

[0099] S5.13, combining equations (2), (30)-(33), the final control law of the lateral vehicle speed adaptive terminal sliding mode controller is obtained:

[0100]

[0101] Where, F yd is the expected total lateral force.

[0102] Preferably, the distributed steer-by-wire vehicle lower layer torque and angle optimization distribution reconstruction controller of step S6 is established by the following steps:

[0103] S6.1. Establish the optimization objective function of the torque optimization distribution reconstruction controller:

[0104]

[0105] Where, F txdij To optimize the distribution of tire longitudinal forces; F tydij To optimize the distribution of tire lateral forces; F zij is the vertical load of the tire; μ represents the road adhesion coefficient; τ ij is the weight coefficient of each wheel, which is used to adjust the proportion of each tire force in the entire optimization objective function;

[0106] S6.2. Establish a second optimization objective function to prevent the first objective function optimization solver from having no solution under complex working conditions and to reduce the error between the longitudinal and lateral forces and the desired target values. The second optimization objective function is as follows:

[0107] min J2=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 ) (36)

[0108] Where η1 and η2 are the longitudinal and lateral force error weight coefficients;

[0109] S6.3. Establish a third optimization objective function to ensure that the optimization objective function has a solution and reduce the error of the yaw moment and the additional yaw moment. The third optimization objective function is as follows:

[0110] minJ3=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 +η3(ΔM-M z ) 2 ) (37)

[0111] Where η3 is the yaw moment error weight coefficient;

[0112] S6.4. Establish an overall optimization model for torque and angle optimization distribution reconstruction controller:

[0113] When a wheel's steering motor fails completely, the output torque of the steering motor is zero, and no lateral force is distributed to the wheel. The optimization distribution controller will optimize the distribution among the remaining steering wheels. The lateral force to be distributed to the faulty wheel is:

[0114]

[0115] Where, failure time ζ ij =0, otherwise ζ ij =1;

[0116] Since the faulty wheel cannot provide the required lateral force, the failure factor is added to the constraint conditions to obtain the overall optimization model:

[0117]

[0118] Where, T ijmax is the maximum output torque of the hub motor;

[0119] S6.5. Establish an inverse model of tire cornering characteristics:

[0120]

[0121] Where, C α is the tire cornering stiffness; p is a constant;

[0122] Combining equations (35)-(40) and (10), the desired lateral forces of the four wheels are obtained, and the desired wheel turning angle δ is obtained by inversely calculating the tire cornering characteristic model. dij .

[0123] Compared with the prior art, the present invention has the following beneficial effects:

[0124] In the design of the upper-layer fault-tolerant controller, the present invention adopts an adaptive terminal sliding mode control method. The adaptive terminal sliding mode control has strong robustness, and the adaptive parameters can be adaptively changed according to changes in the sliding mode surface and the system state. While reducing the sliding mode control chattering, the convergence speed of the sliding mode is improved and the convergence time is reduced. It can take into account the robustness of the system while ensuring the target tracking performance; in the design of the lower-layer torque and angle optimization distribution reconstruction controller, three objective functions are established. On the basis of minimizing the tire load rate, two more objective functions are established to reduce the error between the longitudinal and lateral forces and the desired target values, and the error between the yaw moment and the additional yaw moment, thereby ensuring that the optimization objective function has a solution, so as to improve the handling stability of the distributed wire-controlled steering when the steering fails. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] Figure 1 This is a diagram of the distributed steer-by-wire vehicle dynamics model of the present invention;

[0126] Figure 2 This is a fault-tolerant control flow chart of a distributed steer-by-wire vehicle according to the present invention. DETAILED DESCRIPTION

[0127] The following is combined with Figure 1-2 , the technical solution of the present invention is described in detail.

[0128] like Figure 1 As shown, a distributed steer-by-wire vehicle fault-tolerant control method includes the following steps:

[0129] S1. Establish a distributed steer-by-wire vehicle dynamics model:

[0130] Longitudinal dynamic equation:

[0131]

[0132] Lateral dynamics equation:

[0133]

[0134] Yaw dynamics model:

[0135]

[0136] Among them are:

[0137]

[0138] Where m is the vehicle mass; v x and v y are the components of the vehicle's center of mass velocity on the x-axis and y-axis respectively; a x is the longitudinal acceleration; a y is the lateral acceleration; ω r is the vehicle's yaw rate; β is the vehicle's center of mass sideslip angle; I z is the yaw moment of inertia of the vehicle around the z axis; l f 、l r are the distances from the center of mass of the vehicle to the front and rear axles, respectively; b is half the wheelbase; ij = fl, fr, rl, rr represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively; F tyij 、F txij are tire lateral force and tire longitudinal force respectively; F xij 、F yij F txij 、F tyij The longitudinal force and lateral force of each wheel decomposed and synthesized in the vehicle coordinate system; F x 、F y F txij 、Ftyij The total longitudinal force and total lateral force decomposed and synthesized in the vehicle coordinate system; M z is the yaw moment of the vehicle; δ ij is the wheel angle.

[0139] Establishing the wheel dynamics model:

[0140]

[0141] Where J is the moment of inertia; ω ij is the angular velocity of each wheel; T ij is the driving torque; F zij is the vertical load; f ij is the rolling resistance coefficient; r ij is the rolling radius.

[0142] Considering the vertical load transfer to each wheel caused by the longitudinal and lateral movement of the vehicle during driving, according to the moment balance relationship, the vertical load on each wheel is calculated as follows:

[0143]

[0144] Where g is the acceleration due to gravity and h is the height of the center of mass from the ground.

[0145] Building the "Magic Formula" tire model:

[0146] Y(x)=Dsin{Carctan[Bx-E(Bx-arctan(Bx))]} (7)

[0147] Where Y(x) is the lateral force F of the tire tyij Or the longitudinal force F txij ; x is the tire slip angle α ij Or tire slip rate λ ij ; B, C, D, and E are fitting parameters in the tire characteristic curve, which are stiffness factor, curve shape factor, curve peak factor, and curve curvature factor, respectively.

[0148] The tire slip rate of each wheel is calculated as:

[0149]

[0150] Where u ij is the wheel center speed of each wheel; ω ij is the angular velocity of each wheel; r is the rolling radius of the wheel.

[0151] The wheel center speed of each wheel can be expressed as:

[0152]

[0153] The slip angles of the four wheels are:

[0154]

[0155] S2. Establish a two-degree-of-freedom vehicle reference model:

[0156]

[0157] Where k f is the front axle equivalent cornering stiffness of the reference model; k r is the equivalent cornering stiffness of the rear axle of the reference model; δ is the front wheel turning angle of the reference model.

[0158] The ideal yaw rate of a car in steady-state driving is as follows:

[0159]

[0160] Where, ω rd is the ideal yaw rate of the car when it is driving in a steady state, and K is the stability factor, which can be expressed as:

[0161]

[0162] Considering that the vehicle is affected by the adhesion coefficient between the tires and the ground when driving, the desired yaw rate must satisfy the following formula:

[0163]

[0164] The ideal yaw rate of the reference model based on equations (12) and (14) can be expressed as:

[0165]

[0166] S3. determining a desired longitudinal speed and an ideal lateral speed;

[0167] The longitudinal velocity of the vehicle is adjusted by controlling the longitudinal acceleration. The desired longitudinal velocity can be expressed as:

[0168]

[0169] Where, v xd is the desired longitudinal velocity; v x0 is the longitudinal speed at the initial moment; a xd is the target longitudinal acceleration.

[0170] Let the ideal center of mass sideslip angle be zero, and the ideal lateral speed can be obtained as:

[0171] v yd =v x tanβ d (17)

[0172] Where, v yd is the ideal lateral speed, β d is the ideal center of mass sideslip angle.

[0173] S4. Establish a failure model for the steering motor of a distributed steer-by-wire vehicle:

[0174] The output torque expression of the steering motor fault is:

[0175] u λ =ζu d +Δu (18)

[0176] Where ζ∈[0,1] is the steering motor failure factor; u λ is the fault output torque of the steering motor; u d is the expected output torque of the steering motor; Δu is other faults of the steering motor.

[0177] Considering only the complete failure of a single steering motor wheel with ζ = 0 and Δu = 0, the braking mechanism in the steering system activates, maintaining the wheel angle at the point where the steering motor fails. When a steering motor fails, fault-tolerant control is implemented by applying an additional yaw torque.

[0178] S5. Establish an upper-layer adaptive terminal sliding mode fault-tolerant controller for a distributed steer-by-wire vehicle.

[0179] S6. Establish a distributed steer-by-wire vehicle lower-layer torque and steering angle optimization distribution reconstruction controller.

[0180] like Figure 2 As shown, the establishment of the upper layer adaptive terminal sliding mode fault-tolerant controller in S5 includes the following steps:

[0181] S5.1. Modify the two-degree-of-freedom vehicle reference model to a four-wheel independent steering two-degree-of-freedom vehicle model and add the additional yaw moment ΔM as follows:

[0182]

[0183] S5.2. Establish the tracking error of the additional yaw moment adaptive terminal sliding mode controller:

[0184] e r =ω r -ω rd (20)

[0185] S5.3. Establish the sliding mode surface of the additional yaw moment adaptive terminal sliding mode controller:

[0186]

[0187] Where k is an adaptive parameter, c>0, q and p are positive odd numbers, and the expression of k is:

[0188]

[0189] Wherein, σ>0,ξ>0,0<ε<1; w is an arbitrary real number; e is the natural logarithm.

[0190] S5.4. Use the exponential reaching law to switch the additional yaw moment adaptive terminal sliding mode controller:

[0191]

[0192] Where η>0,κ>0

[0193] S5.5. Combining Equations (19) to (23), we can obtain the final control law of the adaptive terminal sliding mode controller:

[0194]

[0195] Where ΔM is the additional yaw moment, and its positive and negative values ​​represent the vehicle's steering direction, with a positive value indicating a left turn and a negative value indicating a right turn.

[0196] S5.6. Establish a distributed steer-by-wire vehicle longitudinal speed adaptive terminal sliding mode controller. The tracking error is:

[0197] e x =v x -v xd (25)

[0198] S5.7. Establish the sliding mode surface of the longitudinal speed adaptive terminal sliding mode controller for the distributed steer-by-wire vehicle:

[0199]

[0200] Where, c x is the controller parameter, its value is greater than zero, q x 、p x is a positive odd number, k x The expression is:

[0201]

[0202] Where σ x >0,ξ x >0,0<ε x <1;w x is any real number; e is the natural logarithm.

[0203] S5.8. Use the exponential reaching law to switch the longitudinal vehicle speed adaptive terminal sliding mode controller:

[0204]

[0205] Where η x , κ x are controller parameters, and their values ​​are all greater than zero.

[0206] S5.9. Combining equations (1), (25) and (28), the final control law of the longitudinal vehicle speed adaptive terminal sliding mode controller can be obtained:

[0207] Where, F xd is the expected total longitudinal force.

[0208] S5.10. Establish a distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller with a tracking error of:

[0209] e y =v y -v yd (30)

[0210] S5.11. Establish the sliding mode surface of the distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller:

[0211]

[0212] Where, c y is the controller parameter, its value is greater than zero, q y 、p y is a positive odd number, k y The expression is:

[0213]

[0214] Where σ y >0,ξ y >0,0<ε y <1;w y is any real number; e is the natural logarithm.

[0215] S5.12. Use the exponential reaching law to switch the lateral speed adaptive terminal sliding mode controller:

[0216]

[0217] Where η y , κ y are controller parameters, and their values ​​are all greater than zero.

[0218] S5.13, combining equations (2), (30) and (33), we can obtain the final control law of the lateral vehicle speed adaptive terminal sliding mode controller:

[0219]

[0220] Where, F yd is the expected total lateral force.

[0221] like Figure 2 As shown, the distributed steer-by-wire vehicle lower layer torque and angle optimization distribution reconstruction controller in S6 is established by the following steps:

[0222] S6.1. Establish the optimization objective function of the torque optimization distribution reconstruction controller:

[0223]

[0224] Where, F txdij To optimize the distribution of tire longitudinal forces; F tydij To optimize the distribution of tire lateral forces; F zij is the vertical load of the tire; μ represents the road adhesion coefficient; τ ij is the weight coefficient of each wheel, which is used to adjust the proportion of each tire force in the entire optimization objective function.

[0225] S6.2. In order to prevent the first objective function optimization solver from having no solution under complex working conditions and to minimize the error between the longitudinal and lateral forces and the expected target values, a second optimization objective function is established:

[0226] min J2=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 ) (36)

[0227] Where η1 and η2 are the longitudinal and lateral force error weight coefficients.

[0228] S6.3. To further ensure that the optimization objective function has a solution and reduce the error of the yaw moment and the additional yaw moment, a third optimization objective function is established:

[0229] minJ3=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 +η3(ΔM-M z ) 2 ) (37)

[0230] Where η3 is the yaw moment error weight coefficient.

[0231] S6.4. Establish an overall optimization model for torque and angle optimization distribution reconstruction controller:

[0232] When a wheel's steering motor fails completely, the output torque of the steering motor is zero, and no lateral force is distributed to the wheel. The optimization distribution controller will optimize the distribution among the remaining steering wheels. The lateral force to be distributed to the faulty wheel is:

[0233]

[0234] Where, failure time ζ ij =0, otherwise ζ ij =1.

[0235] Since the faulty wheel cannot provide the required lateral force, the failure factor is added to the constraint conditions to obtain the overall optimization model:

[0236]

[0237] Where, T ijmax is the maximum output torque of the hub motor.

[0238] S6.5. Establish an inverse model of tire cornering characteristics:

[0239]

[0240] Where, C α is the tire cornering stiffness; p is a constant.

[0241] Combining equations (35)-(40) and (10), the desired lateral forces of the four wheels can be obtained, and the desired wheel turning angle δ can be obtained by inversely calculating the tire cornering characteristic inverse model. dij .

[0242] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless defined as such, will not be interpreted in an idealized or overly formal sense.

[0243] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A distributed steer-by-wire vehicle fault-tolerant control method, characterized in that: The following steps are involved: S1. Establish a distributed steer-by-wire vehicle dynamics model; S2. Establish a two-degree-of-freedom vehicle reference model; S3. determining a desired longitudinal speed and an ideal lateral speed; S4. Establishing a failure model for the steering motor of a distributed steer-by-wire vehicle; S5. Establishing an upper adaptive terminal sliding mode fault-tolerant controller for a distributed steer-by-wire vehicle; S6. Establishing a distributed steer-by-wire vehicle lower layer torque and steering angle optimization distribution and reconstruction controller; The failure model of the steering motor of a distributed steer-by-wire vehicle is established as follows: The output torque expression of the steering motor fault is: you λ =ζu d +Δu (18) Where ζ∈[0,1] is the steering motor failure factor; u λ is the fault output torque of the steering motor; u d is the expected output torque of the steering motor; Δu is other faults of the steering motor; When ζ = 0 and Δu = 0, the steering motor fails completely, the brake mechanism in the steering system takes effect, and the wheel maintains the wheel angle when the steering motor fails completely, and fault-tolerant control is performed by adding yaw torque; The establishment of the upper layer adaptive terminal sliding mode fault-tolerant controller in S5 includes the following steps: S5.

1. Modify the two-degree-of-freedom vehicle reference model to a four-wheel independent steering two-degree-of-freedom vehicle model and add an additional yaw moment ΔM: S5.

2. Establish the tracking error of the additional yaw moment adaptive terminal sliding mode controller: e r =ω r -oh rd (20) S5.

3. Establish the sliding mode surface of the additional yaw moment adaptive terminal sliding mode controller: Where k is an adaptive parameter, c>0, q and p are positive odd numbers, and the expression of k is: Wherein, σ>0,ξ>0,0<ε<1; w is any real number; e is the natural logarithm; S5.

4. Use the exponential reaching law to switch the additional yaw moment adaptive terminal sliding mode controller: Where η>0,κ>0 S5.

5. Combining Equations (19) to (23), we can obtain the final control law of the adaptive terminal sliding mode controller: Where ΔM is the additional yaw moment, and the positive and negative values ​​represent the vehicle's turning direction, with a positive value indicating a left turn and a negative value indicating a right turn. S5.

6. Establish a distributed steer-by-wire vehicle longitudinal speed adaptive terminal sliding mode controller. The tracking error is: yes x =v x -v xd (25) S5.

7. Establish the sliding mode surface of the longitudinal speed adaptive terminal sliding mode controller for the distributed steer-by-wire vehicle: Where, c x is a controller parameter with a value greater than zero, q x 、p x is a positive odd number, k x The expression is: Where σ x >0,ξ x >0,0<ε x <1;w x is any real number; e is the natural logarithm; S5.

8. Use the exponential reaching law to switch the longitudinal vehicle speed adaptive terminal sliding mode controller: Where η x , κ x are controller parameters with values ​​greater than zero; S5.

9. Combining equations (1), (25) and (28), we can obtain the final control law of the longitudinal vehicle speed adaptive terminal sliding mode controller: Where, F xd is the expected total longitudinal force; S5.

10. Establish a distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller with a tracking error of: yes y =v y -v yd (30) S5.

11. Establish the sliding mode surface of the distributed steer-by-wire vehicle lateral speed adaptive terminal sliding mode controller: Where, c y is a controller parameter with a value greater than zero, q y 、p y is a positive odd number, k y The expression is: Where σ y >0,ξ y >0,0<ε y <1;w y is any real number; e is the natural logarithm; S5.

12. Use the exponential reaching law to switch the lateral speed adaptive terminal sliding mode controller: Where η y , κ y are controller parameters with values ​​greater than zero; S5.13, combining equations (2), (30)-(33), the final control law of the lateral vehicle speed adaptive terminal sliding mode controller is obtained: Where, F yd is the expected total lateral force; The distributed steer-by-wire vehicle lower layer torque and angle optimization distribution reconstruction controller of step S6 is established by the following steps: S6.

1. Establish the optimization objective function of the torque optimization distribution reconstruction controller: Where, F txdij To optimize the distribution of tire longitudinal forces; F tydij To optimize the distribution of tire lateral forces; F zij is the vertical load of the tire; μ represents the road adhesion coefficient; τ ij is the weight coefficient of each wheel, which is used to adjust the proportion of each tire force in the entire optimization objective function; S6.

2. Establish a second optimization objective function to prevent the first objective function optimization solver from having no solution under complex working conditions and to reduce the error between the longitudinal and lateral forces and the desired target values. The second optimization objective function is as follows: min J2=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 ) (36) Where η1 and η2 are the longitudinal and lateral force error weight coefficients; S6.

3. Establish a third optimization objective function to ensure that the optimization objective function has a solution and reduce the error of the yaw moment and the additional yaw moment. The third optimization objective function is as follows: min J3=min(η1(F x -F xd ) 2 +η2(F y -F yd ) 2 +η3(ΔM-M z ) 2 ) (37) Where η3 is the yaw moment error weight coefficient; S6.

4. Establish an overall optimization model for torque and angle optimization distribution reconstruction controller: When a wheel's steering motor fails completely, the output torque of the steering motor is zero, and no lateral force is distributed to the wheel. The optimization distribution controller will optimize the distribution among the remaining steering wheels. The lateral force to be distributed to the faulty wheel is: Where, failure time ζ ij =0, otherwise ζ ij =1; Since the faulty wheel cannot provide the required lateral force, the failure factor is added to the constraint conditions to obtain the overall optimization model: Where, T ijmax is the maximum output torque of the hub motor; S6.

5. Establish an inverse model of tire cornering characteristics: Where, C α is the tire cornering stiffness; p is a constant; Combining equations (35)-(40) and (10), the desired lateral forces of the four wheels are obtained, and the desired wheel turning angle δ is obtained by inversely calculating the tire cornering characteristic model. dij .

2. The distributed steer-by-wire vehicle fault-tolerant control method according to claim 1, characterized in that: The distributed steer-by-wire vehicle dynamics model is established as follows: Longitudinal dynamic equation: Lateral dynamics equation: Yaw dynamics model: Among them are: Where m is the vehicle mass; v x and v y are the components of the vehicle's center of mass velocity on the x-axis and y-axis respectively; a x is the longitudinal acceleration; a y is the lateral acceleration; ω r is the vehicle's yaw rate; β is the vehicle's center of mass sideslip angle; I z is the yaw moment of inertia of the vehicle around the z axis; l f 、l r are the distances from the center of mass of the vehicle to the front and rear axles, respectively; b is half the wheelbase; ij = fl, fr, rl, rr represent the left front wheel, right front wheel, left rear wheel, and right rear wheel of the vehicle, respectively; F tyij 、F txij are tire lateral force and tire longitudinal force respectively; F xij 、F yij F txij 、F tyij The longitudinal force and lateral force of each wheel decomposed and synthesized in the vehicle coordinate system; F x 、F y F txij 、F tyij The total longitudinal force and total lateral force decomposed and synthesized in the vehicle coordinate system; M z is the yaw moment of the vehicle; δ ij is the wheel angle; Establishing the wheel dynamics model: Where J is the moment of inertia; ω ij is the angular velocity of each wheel; T ij is the driving torque; F zij is the vertical load; f ij is the rolling resistance coefficient; r ij is the rolling radius; According to the moment balance relationship, the vertical load on each wheel is calculated as follows: Where g is the acceleration due to gravity; h is the height of the center of mass from the ground; Building the Magic Formula Tire Model: Y(x)=Dsin{Carctan[Bx-E(Bx-arctan(Bx))]} (7) Where Y(x) is the lateral force F of the tire tyij Or the longitudinal force F txij ; x is the tire slip angle α ij Or tire slip rate λ ij ; B, C, D, and E are fitting parameters in the tire characteristic curve, which are stiffness factor, curve shape factor, curve peak factor, and curve curvature factor, respectively.

3. The fault-tolerant control method for a distributed steer-by-wire vehicle according to claim 2, characterized in that: The tire slip rate of each wheel is calculated as: Where u ij is the wheel center speed of each wheel; ω ij is the angular velocity of each wheel; r is the rolling radius of the wheel; The wheel center speed of each wheel can be expressed as:

4. The fault-tolerant control method for a distributed steer-by-wire vehicle according to claim 2, characterized in that: The slip angles of the four wheels are calculated as:

5. The distributed steer-by-wire vehicle fault-tolerant control method according to claim 1, characterized in that: The establishment of the two-degree-of-freedom vehicle reference model is specifically as follows: Where k f is the front axle equivalent cornering stiffness of the reference model; k r is the equivalent cornering stiffness of the rear axle of the reference model; δ is the front wheel turning angle of the reference model; The ideal yaw rate of a car in steady-state driving is as follows: Where, ω rd is the ideal yaw rate of the car when it is driving in a steady state, and K is the stability factor: The expected yaw rate must satisfy the following formula: Where μ represents the road adhesion coefficient; the ideal yaw rate of the reference model based on equations (12) and (14) can be expressed as:

6. The distributed steer-by-wire vehicle fault-tolerant control method according to claim 1, characterized in that: The desired longitudinal velocity is expressed as: Where, v xd is the desired longitudinal velocity; v x0 is the longitudinal speed at the initial moment; a xd is the target longitudinal acceleration.

7. The fault-tolerant control method for a distributed steer-by-wire vehicle according to claim 1, characterized in that: The ideal lateral vehicle speed is expressed as: v yd =v x ·tanβ d (17) Where, v yd is the ideal lateral speed, β d is the ideal center of mass sideslip angle.

Images