Fault-tolerant control method and device for actuators of electric vehicle, storage medium and vehicle

Abstract

The application discloses an actuator fault-tolerant control method and device of an electric vehicle, a storage medium and the electric vehicle, and belongs to the technical field of electric vehicles. The method comprises the following steps: acquiring vehicle sensor information and a reference trajectory of a distributed drive electric vehicle; performing trajectory tracking on the vehicle sensor information and the reference trajectory by using an MPC algorithm to obtain an expected front wheel rotation angle; when it is determined that at least one steering actuator fails according to the vehicle sensor information, performing calculation on the expected front wheel rotation angle and an actual front wheel rotation angle at the last moment by using a sliding mode control (SMC) algorithm to obtain a differential torque; when it is determined that at least one drive actuator fails according to the vehicle sensor information, performing calculation on the differential torque by using an optimal distribution algorithm based on tire adhesion utilization minimization to obtain a current moment; and controlling the steering of the distributed drive electric vehicle according to the current moment. The application realizes reliable control of the vehicle under the condition that the steering actuator and the drive actuator fail simultaneously.

Description

Technical Field

[0001] This application relates to the field of electric vehicle technology, and in particular to an actuator fault-tolerant control method, device, storage medium, and vehicle for an electric vehicle. Background Technology

[0002] With increasing environmental awareness and a worsening energy crisis, electric vehicles (EVs), as clean and efficient modes of transportation, have gained widespread attention and popularity. Distributed-drive electric vehicles (DDEVs) possess advantages such as short drive chains, high transmission efficiency, fast motor response, and precise torque control, making them one of the ideal carriers for electric vehicles. However, as an overdrive system, DDEVs exhibit high system complexity, strong coupling between subsystems, and a relatively large number of actuators and sensors, significantly increasing the probability of subsystem failures. Therefore, researching fault-tolerant control for DDEVs is crucial for improving overall vehicle safety and driving stability.

[0003] Depending on the actuators used in distributed electric vehicles, fault-tolerant control can be divided into fault-tolerant control for steering-by-wire failure and fault-tolerant control for drive motor failure. Existing technologies address the steering-by-wire failure problem by proposing a hierarchical collaborative fault-tolerant control method that integrates differential steering and yaw moment. Based on the upper-level controller solving for the desired front wheel steering angle and additional yaw moment, the lower-level controller optimizes the tire load ratio using the effective set method to achieve output torque distribution.

[0004] Fault-tolerant control for drive motor failure mainly refers to the control method that, when one or more drive motors in a drive-by-wire chassis drive system fail, appropriately redistributes the drive to the remaining drive motors, enabling the vehicle to maintain driving stability and safety. Fault-tolerant control for drive motor failure can generally be divided into passive fault-tolerant control and active fault-tolerant control. Passive fault-tolerant control focuses primarily on drive motor failure control, typically targeting specific fault modes. That is, when the vehicle experiences a corresponding fault, it is not necessary to know the system fault information; only pre-defined logical rules are needed to make the rule-based control strategy robust to faults. Existing technologies have studied the path tracking problem of electric vehicles with a four-electromechanical wheel system under normal and fault conditions. Considering wheel slip constraints and certain actuator faults, a passive fault-tolerant controller based on variable structure control was developed to maintain system stability and ensure acceptable tracking performance. However, for overdrive systems with multiple fault sources, the passive fault-tolerant control strategy can only handle a limited number of fault types, making it difficult to achieve optimal fault-tolerant control performance. Active fault-tolerant control can adopt more appropriate control strategies based on different fault occurrence conditions, adjust the fault-tolerant controller parameters in real time, and actively eliminate or mitigate motor faults, minimizing the impact of faults on vehicle driving status and ensuring overall vehicle stability and safety. Existing technologies address the tracking control problem of four-wheel independent drive vehicles under actuator faults and disturbances by designing a composite observer to simultaneously estimate the state and disturbances of the drive system, reconstructing the fault efficiency factor, and adaptively adjusting controller parameters, thereby improving the robustness of tracking performance.

[0005] The above studies primarily focus on vehicle dynamics and stability as control objectives, addressing fault-tolerant control in distributed electric vehicles in cases of sensor failure, steering actuator failure, and drive actuator failure. However, existing research mainly considers the failure of a single actuator, lacking research on the coupling mechanism of multiple actuator failures and related fault-tolerant control. When multiple actuators fail simultaneously, applying fault-tolerant control only to a single actuator will lead to control failure due to the coupling between actuators, thus seriously threatening driving safety. Summary of the Invention

[0006] This application provides a fault-tolerant control method, device, storage medium, and vehicle for electric vehicles, addressing the problem of inability to perform fault-tolerant control when multiple actuators fail simultaneously, threatening driving safety. The technical solution is as follows:

[0007] A first aspect of this application provides an actuator fault-tolerant control method for an electric vehicle, used in a distributed drive electric vehicle employing front-wheel drive steering, the method comprising:

[0008] Acquire the onboard sensor information and reference trajectory of the distributed drive electric vehicle;

[0009] The Model Predictive Control (MPC) algorithm is used to track the onboard sensor information and the reference trajectory to obtain the desired front wheel steering angle.

[0010] When at least one steering actuator malfunction is determined based on the information from the on-board sensors, the sliding mode control (SMC) algorithm is used to calculate the desired front wheel steering angle and the actual front wheel steering angle at the previous moment to obtain the differential torque.

[0011] When at least one drive actuator malfunction is determined based on the vehicle sensor information, the differential torque is calculated based on an optimized allocation algorithm that minimizes tire adhesion utilization to obtain the current torque of each wheel.

[0012] The steering of the distributed drive electric vehicle is controlled based on the torque at the current moment.

[0013] In one possible implementation, the step of using the MPC algorithm to track the vehicle sensor information and the reference trajectory to obtain the desired front wheel steering angle includes:

[0014] Based on the vehicle sensor information and the reference trajectory, an MPC trajectory tracking optimization objective function is constructed.

[0015] The objective function for optimizing MPC trajectory tracking is transformed into a quadratic function;

[0016] The quadratic function is solved using a quadratic programming method and the first constraint condition to obtain the optimal control increment sequence in the control time domain;

[0017] The first n control variables in the optimal control increment sequence are determined as the desired front wheel steering angle, where n is a positive integer.

[0018] In one possible implementation, the step of creating the MPC trajectory tracking optimization objective function based on the onboard sensor information and the reference trajectory includes:

[0019] Based on the vehicle sensor information and the reference trajectory, the vehicle dynamics model equation and the trajectory tracking model equation are combined to obtain the state-space expression;

[0020] Discretize the parameters in the state-space expression to obtain the discretized gain matrix equation;

[0021] Substituting the discretized gain matrix equation into the state-space expression yields the MPC trajectory tracking optimization objective function.

[0022] In one possible implementation, transforming the MPC trajectory tracking optimization objective function into a quadratic function includes:

[0023] Predict the future state in the time domain based on the MPC trajectory tracking optimization objective function;

[0024] Generate a quadratic function based on the future state in the time domain.

[0025] In one possible implementation, the step of using the SMC algorithm to calculate the differential torque based on the desired front wheel steering angle and the actual front wheel steering angle at the previous moment includes:

[0026] Construct a dynamic model of the steering actuator, wherein the dynamic model includes unknown differential torque;

[0027] Construct a sliding surface based on the desired front wheel steering angle and the actual front wheel steering angle;

[0028] Taking the first-order differential of the sliding surface, we obtain the differential equation;

[0029] The differential torque is calculated based on the dynamic model and the differential equation.

[0030] In one possible implementation, the optimization allocation algorithm based on minimizing tire adhesion utilization calculates the differential torque to obtain the current torque of each wheel, including:

[0031] Obtain fault information of the drive actuator, and construct the failure mode based on the fault information;

[0032] Based on the failure modes, an optimization objective function based on minimizing tire adhesion utilization is constructed;

[0033] The objective function is solved using a quadratic programming method and a second constraint to obtain the current torque of each wheel.

[0034] In one possible implementation, the second constraint is created based on the failure mode, the total driving torque requirement, the differential torque requirement, and the requirement that no additional yaw moment occurs in the effective solution.

[0035] A second aspect of this application provides an actuator fault-tolerant control device for an electric vehicle in a distributed drive electric vehicle using front-wheel drive steering, the device comprising:

[0036] The acquisition module is used to acquire the on-board sensor information and reference trajectory of the distributed drive electric vehicle;

[0037] The tracking module is used to track the vehicle sensor information and the reference trajectory using the Model Predictive Control (MPC) algorithm to obtain the desired front wheel steering angle;

[0038] The calculation module is used to calculate the differential torque by using the sliding mode control (SMC) algorithm to calculate the desired front wheel steering angle and the actual front wheel steering angle at the previous moment when at least one steering actuator failure is determined based on the vehicle sensor information.

[0039] The calculation module is also used to calculate the differential torque based on an optimized allocation algorithm that minimizes tire adhesion utilization when at least one drive actuator is determined to be faulty according to the vehicle sensor information, so as to obtain the current torque of each wheel.

[0040] A control module is used to control the steering of the distributed drive electric vehicle based on the torque at the current moment.

[0041] A third aspect of this application provides a computer-readable storage medium storing at least one instruction, which is loaded and executed by a processor to implement the actuator fault-tolerant control method for an electric vehicle as described above.

[0042] A fourth aspect of this application provides a distributed drive electric vehicle, the distributed drive electric vehicle including a processor and a memory, the memory storing at least one instruction, the instruction being loaded and executed by the processor to implement the actuator fault-tolerant control method of the electric vehicle as described above.

[0043] The beneficial effects of the technical solution provided in this application include at least the following:

[0044] When at least one steering actuator fails, causing the vehicle to completely or partially lose its steering ability, the differential torque can be calculated based on the actual front wheel steering angle and the calculated desired front wheel steering angle. When at least one drive actuator fails, the driving torque of each wheel can be reasonably distributed based on the differential torque, thereby enabling the distributed drive electric vehicle to have better tracking performance. The fault-tolerant control method of the hierarchical architecture can effectively improve the fault tolerance and robustness of the steering system, realize reliable control of the vehicle when both the steering actuator and the drive actuator fail simultaneously, and improve driving safety. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart of an actuator fault-tolerant control method for an electric vehicle provided in one embodiment of this application;

[0047] Figure 2 This is a flowchart of an actuator fault-tolerant control method for an electric vehicle provided in one embodiment of this application;

[0048] Figure 3 This is a schematic diagram of the planar motion of a vehicle provided in one embodiment of this application;

[0049] Figure 4 This is a schematic diagram of the differential steering principle provided in one embodiment of this application;

[0050] Figure 5 This is a schematic diagram of an actuator fault-tolerant control method for a distributed drive electric vehicle provided in one embodiment of this application;

[0051] Figure 6 This is a trajectory diagram under uniform acceleration linear driving conditions provided in one embodiment of this application;

[0052] Figure 7 This is a longitudinal velocity diagram under uniform acceleration linear conditions provided in one embodiment of this application;

[0053] Figure 8 This is an experimental result diagram of a step steering condition provided in one embodiment of this application;

[0054] Figure 9 This is an experimental result diagram under the dual-line-shifting condition provided in one embodiment of this application;

[0055] Figure 10 This is a torque distribution diagram provided in one embodiment of this application;

[0056] Figure 11 This is a structural block diagram of an actuator fault-tolerant control device for an electric vehicle provided in one embodiment of this application. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0058] like Figure 1 The diagram illustrates a flowchart of an actuator fault-tolerant control method for an electric vehicle according to an embodiment of this application. This actuator fault-tolerant control method can be applied to distributed drive electric vehicles using front-wheel drive steering. The actuator fault-tolerant control method for the electric vehicle may include:

[0059] Step 101: Obtain onboard sensor information and reference trajectory of the distributed drive electric vehicle.

[0060] Distributed drive electric vehicles are equipped with multiple on-board sensors, which collect information to obtain on-board sensor information.

[0061] The vehicle-mounted sensor information includes, but is not limited to, the following: vehicle longitudinal speed, vehicle lateral speed, vehicle yaw rate, vehicle front and rear axle lateral stiffness, front wheel steering angle, vehicle center of gravity sideslip angle, vehicle moment of inertia, lateral error between the vehicle center of gravity and the reference trajectory, and yaw angle error between the vehicle heading angle and the trajectory centerline.

[0062] The reference trajectory is a pre-set trajectory, and the distributed drive electric vehicle needs to drive according to the reference trajectory.

[0063] Step 102: Use the MPC algorithm to track the vehicle sensor information and reference trajectory to obtain the desired front wheel steering angle.

[0064] The principle of MPC (Model-based Predictive Control) algorithms is that in each computational step, the system receives state information from the external environment and predicts possible future states through a state update function. Simultaneously, given a reference state, the algorithm is optimized to minimize a pre-defined performance index, thus obtaining the optimal control increment sequence. Typically, the first or first few control increments are selected as the control output. This optimization employs a rolling optimization approach, continuously refining the algorithm until the entire task is completed.

[0065] Step 103: When at least one steering actuator is determined to be faulty based on the information from the on-board sensors, the SMC algorithm is used to calculate the expected front wheel steering angle and the actual front wheel steering angle at the previous moment to obtain the differential torque.

[0066] The core idea of ​​the SMC (Sliding Mode Control) algorithm is to establish a sliding surface, pull the system onto the sliding surface, and make the system move along the sliding surface.

[0067] The actual front wheel steering angle at the previous moment can be obtained from the information from the vehicle's onboard sensors.

[0068] Step 104: When at least one drive actuator is determined to be faulty based on the information from the vehicle-mounted sensors, the differential torque is calculated based on the optimization allocation algorithm that minimizes tire adhesion utilization, and the current torque of each wheel is obtained.

[0069] Since the vehicle's handling stability is better when the sum of the tire adhesion utilization rates is minimized, we can set an optimization allocation algorithm based on the goal of minimizing tire adhesion utilization rates, and rationally allocate the driving torque of each wheel through the optimization allocation algorithm.

[0070] Step 105: Control the steering of the distributed drive electric vehicle based on the torque at the current moment.

[0071] In summary, the actuator fault-tolerant control method for electric vehicles provided in this application can calculate the differential torque based on the actual front wheel angle and the calculated desired front wheel angle when at least one steering actuator fails, causing the vehicle to completely or partially lose its steering ability. When at least one drive actuator fails, the driving torque of each wheel can be reasonably distributed based on the differential torque, thereby enabling the distributed drive electric vehicle to have better tracking performance. The fault-tolerant control method with a hierarchical architecture can effectively improve the fault tolerance and robustness of the steering system, realize reliable control of the vehicle when both the steering actuator and the drive actuator fail simultaneously, and improve driving safety.

[0072] like Figure 2 The diagram illustrates a flowchart of an actuator fault-tolerant control method for an electric vehicle according to an embodiment of this application. This actuator fault-tolerant control method can be applied to distributed drive electric vehicles using front-wheel drive steering. The actuator fault-tolerant control method for the electric vehicle may include:

[0073] Step 201: Obtain onboard sensor information and reference trajectory of the distributed drive electric vehicle.

[0074] Distributed drive electric vehicles are equipped with multiple on-board sensors, which collect information to obtain on-board sensor information.

[0075] The vehicle-mounted sensor information includes, but is not limited to, the following: vehicle longitudinal speed, vehicle lateral speed, vehicle yaw rate, vehicle front and rear axle lateral stiffness, front wheel steering angle, vehicle center of gravity sideslip angle, vehicle moment of inertia, lateral error between the vehicle center of gravity and the reference trajectory, and yaw angle error between the vehicle heading angle and the trajectory centerline.

[0076] The reference trajectory is a pre-set trajectory, and the distributed drive electric vehicle needs to drive according to the reference trajectory.

[0077] Step 202: Construct the MPC trajectory tracking optimization objective function based on the vehicle sensor information and reference trajectory; transform the MPC trajectory tracking optimization objective function into a quadratic function; solve the quadratic function using a quadratic programming method and the first constraint condition to obtain the optimal control increment sequence in the control time domain; determine the first n control quantities in the optimal control increment sequence as the desired front wheel steering angle, where n is a positive integer.

[0078] Specifically, creating an MPC trajectory tracking optimization objective function based on onboard sensor information and a reference trajectory can include the following steps:

[0079] (1) Based on the vehicle sensor information and the reference trajectory, the vehicle dynamics model equation and the trajectory tracking model equation are combined to obtain the state space expression.

[0080] Figure 3 A schematic diagram of the vehicle's planar motion is shown, based on... Figure 3 The parameters shown can be used to establish vehicle dynamics models and trajectory tracking models.

[0081] A. Establishing a vehicle dynamics model

[0082] To solve the trajectory tracking problem in multi-actuator failure-tolerant control of distributed drive electric vehicles, it is necessary to construct a control-oriented vehicle dynamics model that ignores pitch and roll, considers only front-wheel steering, and has a hub motor on each wheel hub.

[0083]

[0084] Where m is the vehicle mass, v x v is the longitudinal velocity of the vehicle. y Let C be the vehicle's lateral velocity, γ be the vehicle's yaw rate, and C be the lateral velocity. f and C r δ represents the lateral stiffness of the front and rear axles of the vehicle. f β is the front wheel steering angle of the vehicle, β is the sideslip angle of the vehicle's center of gravity, and l f and l r I is the distance from the front and rear axles of the vehicle to the center of gravity. z These parameters, such as the vehicle's moment of inertia, can be obtained from information from onboard sensors.

[0085] B. Establish a trajectory tracking model

[0086]

[0087] Among them, e y ψ is the lateral error between the vehicle's center of mass and the target trajectory. e Let ψ be the heading angle error between the vehicle's heading angle and the trajectory centerline, and ρ be the curvature of the reference trajectory. e The angles of both ψ and β are very small and approximately 0, which leads to the conclusion that cos(ψ) e )≈1,sin(ψ) e )≈ψ e And tan(β)≈β. Therefore, based on β≈v y / v x A trajectory tracking model can be obtained.

[0088] C. Establishing the state-space representation

[0089] To achieve superior control performance, this embodiment uses the MPC algorithm to design the upper-level trajectory tracking controller. The principle of the MPC algorithm is that in each calculation step, the system obtains state information from the external environment and predicts possible future states through a state update function. Simultaneously, under a given reference state, the algorithm is optimized to minimize pre-set performance indicators, thereby obtaining the optimal control increment sequence. Generally, the first or first few control increments are selected as the control output. This optimization employs a rolling optimization approach, continuously refining the algorithm until the entire algorithm task is completed.

[0090] By combining the vehicle dynamics model and the trajectory tracking model, the state-space expression can be obtained, as shown below:

[0091]

[0092] in,

[0093] Choose x = [β γ e] y ψ e [δ] represents the vehicle system state variables, namely, the vehicle's sideslip angle, yaw rate, lateral error between the vehicle's center of gravity and the target trajectory, and yaw angle error between the vehicle's heading angle and the trajectory centerline, respectively. f [ ] is the control variable, representing the front wheel steering angle of the vehicle, w = ρv x This is due to external interference.

[0094] (2) Discretize the parameters in the state-space expression to obtain the discretized gain matrix equation.

[0095] Discretization using the zero-order hold technique yields the discretized gain matrix equation:

[0096]

[0097] Among them, A d B d D d Let A, B, and D represent the discretization gain matrices, respectively, and T be the discretization step size.

[0098] (3) Substitute the discretized gain matrix equation into the state space expression to obtain the MPC trajectory tracking optimization objective function.

[0099] Specifically, transforming the MPC trajectory tracking optimization objective function into a quadratic function can include the following steps:

[0100] (1) Predict the future state in the time domain based on the objective function of MPC trajectory tracking optimization.

[0101] Based on the state-space expression equation, the prediction of future states in the time domain can be derived as follows:

[0102] Y = S x x(t)+S u u(t-1)+S Δ ΔU+S w W (5)

[0103] in, Np is the prediction time-domain step size of the trajectory tracking controller, and Nc is the pre-control time-domain step size of the trajectory tracking controller.

[0104] n is the dimension of the state variables in the state space, and m is the dimension of the control variables in the state space.

[0105] (2) Generate a quadratic function based on the future state in the time domain.

[0106] After obtaining the future state Y in the predicted time domain, the control increment ΔU (desired front wheel steering angle) of the system at the current moment can be obtained by solving the constrained optimization objective function. In order to track the reference trajectory in real time, the optimization objective is to minimize the error between the system output and the desired value. Considering the mechanical structure limitations of the vehicle steering actuator and the risk of loss of vehicle control due to excessive changes in the front wheel steering angle, the control increment ΔU should be as small as possible.

[0107] To satisfy the constraints of the mechanical structure and motion conditions, constraints must be imposed on the objective function and some states of the system to achieve the best control effect. In other words, Δδ must be ensured. f δ f β, γ, α f α r Satisfying the corresponding constraints is necessary to achieve lateral motion control of the vehicle and the reliability of the linear model. The first constraint conditions for β and γ can be obtained based on the steady-state conditions:

[0108]

[0109] In this case, it is assumed that μ in formula (6) is known.

[0110] In summary, the performance index J of the trajectory tracking controller can be... mpc Defined as:

[0111]

[0112] in, For systematic error (Y) r The weight matrix of (t)-Y(t)), To control the weight matrix of the increment (ΔU(t)), a relaxation factor ε is added to avoid situations where the solution cannot be obtained. ρ is the weight coefficient of the relaxation factor. The reference system output is a time-varying variable determined by the reference trajectory and system state at each moment, in Y. r China e y Reference value e yr and ψ e Reference value ψ er It should be 0, the reference value for β is β. r and the reference value of γ r for:

[0113]

[0114] in,

[0115] Given the current state variable y(t) and the control variable u(t) from the previous time step, we can use quadratic programming to optimize the quadratic function within the control cycle, i.e., Equation 7, to obtain a series of optimal control increment sequences in the control time domain:

[0116]

[0117] Then, by applying the first n control increments in the control increment sequence as actual control input increments to the system, the desired front wheel steering angle δ at the current moment can be obtained. f , where n is a positive integer. In one embodiment, we can use the first control increment in the control increment sequence as the desired front wheel angle.

[0118] In a distributed drive electric vehicle, if all steering actuators are functioning normally, the desired front wheel steering angle can be directly applied to the steering actuators to control them and achieve trajectory tracking. However, if a steering actuator fails, the current differential torque needs to be determined based on the desired front wheel steering angle, and torque distribution needs to be performed based on this differential torque.

[0119] Step 203: When at least one steering actuator is determined to be faulty based on the information from the vehicle-mounted sensors, the SMC algorithm is used to calculate the expected front wheel steering angle and the actual front wheel steering angle at the previous moment to obtain the differential torque.

[0120] Differential steering principle as follows Figure 4 As shown, the longitudinal driving force of the left and right wheels of the front axle will generate a torque τ around the kingpin. σ and τ r When the torque τ l and τ r When a difference exists, a differential torque is generated, which can be controlled by the return torque τ of the left and right wheels.α To achieve differential steering, the differential torque can be expressed as:

[0121]

[0122] in, The kingpin inclination angle is σ, and the kingpin inclination angle is σ'.

[0123] The following sliding mode front wheel steering angle tracking control strategy is used to determine the current differential torque.

[0124] Specifically, the SMC algorithm is used to calculate the differential torque by combining the desired front wheel steering angle with the actual front wheel steering angle at the previous moment. This process can include the following steps:

[0125] (1) Construct a dynamic model of the steering actuator, which includes unknown differential torque.

[0126] Considering that the frictional resistance torque of the steering actuator and the second derivative of the front wheel steering angle are usually small, they can be ignored as external disturbances. Therefore, the dynamic model of the steering actuator can be simplified to:

[0127]

[0128] Among them, b e For the equivalent damping of the steering actuator, τ α This is the sum of the self-aligning torques of the left and right wheels around their respective kingpins. Self-aligning torque τ α The self-aligning torque is generated by the lateral force of the tire. and the tire's own self-centering torque M zz =M z ×cosζcosσ and the restoring torque generated by the vertical load on the tire. Composed of various elements, including F. y M is the lateral force of the tire. z For the tire's self-aligning torque, F z This refers to the vertical load on the tire.

[0129] (2) Construct a sliding surface based on the desired front wheel angle and the actual front wheel angle.

[0130] In order to obtain the front wheel steering angle command δ from the trajectory tracking controller fd To achieve effective tracking, construct a sliding surface as follows:

[0131]

[0132] (3) Perform a first-order differential on the sliding surface to obtain the differential equation.

[0133] Taking the first-order differential of the synovial surface, we get:

[0134]

[0135] (4) Calculate the differential torque based on the dynamic model and differential equation.

[0136] According to formula (12), we can obtain:

[0137] ΔM f =b e u-τ α (13)

[0138] in, Set error value x i The damping ratio is Δ, the sliding surface thickness is w. n =|1 / x i Δ| represents the natural frequency.

[0139] Step 204: When at least one drive actuator is determined to be faulty based on the information from the vehicle-mounted sensors, the differential torque is calculated based on the optimization allocation algorithm that minimizes tire adhesion utilization, and the current torque of each wheel is obtained.

[0140] Since the vehicle's handling stability is better when the sum of the tire adhesion utilization rates is minimized, we can set an optimization allocation algorithm based on the goal of minimizing tire adhesion utilization rates, and rationally allocate the driving torque of each wheel through the optimization allocation algorithm.

[0141] After obtaining the differential torque, a lower-level torque distribution controller can be established, which will then distribute the torque.

[0142] Specifically, the differential torque is calculated based on an optimization allocation algorithm that minimizes tire adhesion utilization, to obtain the current torque of each wheel. This process may include the following steps:

[0143] (1) Obtain fault information of the drive actuator and construct the failure mode based on the fault information.

[0144] The failure mode of the actuator is defined as:

[0145] T i '=m i T i

[0146]

[0147] Among them, T i 'T' represents the actual output torque of the drive motor. i For the desired output torque of the drive motor, m i The fault diagnosis factor is output by the drive system fault diagnostic tool.

[0148] (2) Construct an optimization objective function based on minimizing tire adhesion utilization rate according to the failure mode.

[0149] Establish an optimization objective function based on minimizing tire adhesion utilization:

[0150]

[0151] Among them, F xi F is the longitudinal force on each wheel of the vehicle. zi This refers to the vertical load on each wheel of the vehicle.

[0152] (3) The objective function is solved by using the quadratic programming method and the second constraint to obtain the current torque of each wheel.

[0153] In the process of optimizing the allocation algorithm, it is necessary not only to ensure that the effective solution can meet the requirements of the total longitudinal driving torque and differential torque, but also to ensure that the effective solution does not produce additional yaw torque, and to consider the influence of factors such as the maximum output torque of the drive motor and fault information. In other words, the second constraint condition is created based on the failure mode, the total driving torque requirement, the differential torque requirement, and the requirement that the effective solution does not produce additional yaw torque. Therefore, the constraint condition is:

[0154]

[0155] The quadratic programming method is used to solve for the driving torque of each wheel, and the resulting performance index is:

[0156]

[0157] in, And ζ=diag{ζ1,ζ2,ζ3} is the weight matrix that satisfies the equality condition, where ζ1 is the weight ΔM of the differential moment. f ζ2 represents the required longitudinal torque T tot The weights are ζ3 and ζ3, which are the weights of the additional yaw moment generated.

[0158] u = [T fl T fr T rl T rr [] is the solver's optimization vector, representing the output torque of the hub motors at all four wheels of the vehicle. v = [ΔM] f T tot 0] T Differential torque ΔM f Demand for longitudinal torque T tot And the expected value of the additional yaw moment (which is set to 0 to avoid affecting the trajectory tracking process), where K is the gain matrix.

[0159] Step 205: Control the steering of the distributed drive electric vehicle based on the torque at the current moment.

[0160] In summary, the actuator fault-tolerant control method for electric vehicles provided in this application can calculate the differential torque based on the actual front wheel angle and the calculated desired front wheel angle when at least one steering actuator fails, causing the vehicle to completely or partially lose its steering ability. When at least one drive actuator fails, the driving torque of each wheel can be reasonably distributed based on the differential torque, thereby enabling the distributed drive electric vehicle to have better tracking performance. The fault-tolerant control method with a hierarchical architecture can effectively improve the fault tolerance and robustness of the steering system, realize reliable control of the vehicle when both the steering actuator and the drive actuator fail simultaneously, and improve driving safety.

[0161] In this embodiment, the distributed drive electric vehicle is equipped with a trajectory tracking controller, a steering failure controller, a steering system fault detector, a torque distribution controller, a drive system fault detector, and a vehicle speed tracker, such as... Figure 5 As shown below, the process of the fault-tolerant control method will be explained in conjunction with the structure of a distributed drive electric vehicle.

[0162] 1. The trajectory tracking controller generates the desired front wheel steering angle based on onboard sensor information and a reference trajectory;

[0163] 2. The steering system fault detector detects whether there is a fault in the steering actuator based on the information from the vehicle's onboard sensors. If there is a fault in the steering actuator, the steering failure controller is triggered to work.

[0164] 3. The steering failure controller generates differential torque based on the desired front wheel steering angle and the actual front wheel steering angle at the previous moment;

[0165] 4. The drive system fault detector detects whether there is a fault in the drive actuator based on the information of the vehicle sensor. If there is a fault in the drive actuator, the torque distribution controller is triggered to work and the fault parameters are sent to the torque distribution controller.

[0166] 5. The vehicle speed tracker generates the longitudinal demand torque based on the target vehicle speed and the actual vehicle speed;

[0167] 6. The torque distribution controller generates the current torque of each wheel based on the differential torque, fault parameters, and longitudinal demand torque, and controls the steering of the distributed drive electric vehicle according to the current torque.

[0168] The following is a simulation experiment analysis of the above fault-tolerant control method:

[0169] To verify the effectiveness of the above fault-tolerant control scheme, we conducted experiments under three conditions: uniform acceleration linear operation, step steering operation, and double lane change operation. The experimental platform was jointly built by MATLAB / Simulink and Carsim, and the vehicle parameters used in the experiment are shown in Table 1.

[0170] 1. Uniformly accelerated linear driving condition

[0171] A simulation experiment was conducted on a vehicle accelerating linearly at speeds ranging from 30 km / h to 60 km / h. The experiment was set with failure rates of the vehicle's left front wheel drive motor at 100%, 75%, 50%, and 25%. Figure 6 As can be seen from the trajectory diagram, as the failure rate increases, vehicles without fault-tolerant control will deviate more from the trajectory, while vehicles with fault-tolerant controllers will always maintain straight-line travel. Figure 7 The longitudinal speed results show that the greater the failure rate, the worse the speed tracking performance of vehicles without fault-tolerant control, while vehicles with fault-tolerant controllers always have good speed tracking performance.

[0172] 2. Step turning condition

[0173] A step steering simulation experiment was conducted, in which a step steering angle of 0.1 rad was generated at 1 second, and a steering actuator failure was simulated at 3 seconds. The experimental results are as follows. Figure 8 As shown in the figure, when the steering actuator fails at 3 seconds, the vehicle without fault-tolerant control loses its steering ability, the yaw rate drops to 0, the yaw angle remains unchanged, and the vehicle continues to travel along the current heading angle; while the experimental results of the vehicle with the fault-tolerant controller after the failure are basically the same as those without the failure. This result shows that the steering failure fault-tolerant controller proposed in this paper has good fault-tolerant performance.

[0174] 3. Double line shifting operation

[0175] The dual lane change test was conducted under the following conditions: vehicle speed 60 km / h, road adhesion coefficient 0.85. Three types of vehicles were used, and the designed fault-tolerant control method was implemented. Vehicle A was a normal vehicle, vehicle B was a vehicle with only steering actuator failure, and vehicle C was a vehicle with both steering actuator and drive motor completely failed. The experimental results are as follows: Figure 9 As shown, vehicles B and C lost their steering ability due to steering actuator failure, but all three vehicles maintained good trajectory tracking accuracy and stability, demonstrating the effectiveness of the fault-tolerant control method presented in this paper for steering failure. Figure 10As shown in the torque distribution diagram, vehicle B exhibits a different torque distribution, requiring the controller to generate differential torque to compensate for steering actuator failure. Vehicle C suffers from steering actuator and left front wheel drive motor failures, but the lower-level torque distribution controller can immediately allocate torque to the remaining three drive motors to meet the vehicle trajectory tracking control requirements. In summary, the experimental results demonstrate that the proposed multi-actuator failure fault-tolerant control method possesses excellent fault-tolerant performance and can improve the safety and reliability of distributed drive electric vehicles.

[0176] Table 1 Vehicle Parameters

[0177]

[0178] like Figure 11 The diagram illustrates a structural block diagram of an actuator fault-tolerant control device for an electric vehicle according to an embodiment of this application. This actuator fault-tolerant control device can be applied to a distributed drive electric vehicle using front-wheel drive steering. The actuator fault-tolerant control device for the electric vehicle may include:

[0179] The acquisition module 1110 is used to acquire onboard sensor information and reference trajectory of the distributed drive electric vehicle;

[0180] Tracking module 1120 is used to track the vehicle sensor information and reference trajectory using the model predictive control (MPC) algorithm to obtain the desired front wheel steering angle;

[0181] The calculation module 1130 is used to calculate the differential torque by using the sliding mode control (SMC) algorithm to calculate the desired front wheel steering angle and the actual front wheel steering angle at the previous moment when at least one steering actuator failure is determined based on the information from the on-board sensors.

[0182] The calculation module 1130 is also used to calculate the differential torque based on an optimized allocation algorithm that minimizes tire adhesion utilization when at least one drive actuator failure is determined based on on-board sensor information, so as to obtain the current torque of each wheel.

[0183] The control module 1140 is used to control the steering of the distributed drive electric vehicle based on the torque at the current moment.

[0184] In an optional embodiment, the tracking module 1120 is further configured to:

[0185] Construct an MPC trajectory tracking optimization objective function based on vehicle sensor information and reference trajectory;

[0186] The objective function for MPC trajectory tracking optimization is transformed into a quadratic function;

[0187] The quadratic function is solved using a quadratic programming method and the first constraint condition to obtain the optimal control increment sequence in the control time domain;

[0188] The first n control variables in the optimal control increment sequence are determined as the desired front wheel steering angle, where n is a positive integer.

[0189] In an optional embodiment, the tracking module 1120 is further configured to:

[0190] Based on the information from the onboard sensors and the reference trajectory, the equations of the vehicle dynamics model and the trajectory tracking model are combined to obtain the state-space expression;

[0191] Discretize the parameters in the state-space expression to obtain the discretized gain matrix equation;

[0192] Substituting the discretized gain matrix equation into the state-space expression, we obtain the MPC trajectory tracking optimization objective function.

[0193] In an optional embodiment, the tracking module 1120 is further configured to:

[0194] Predict future states in the time domain by optimizing the objective function based on MPC trajectory tracking;

[0195] Generate a quadratic function based on the future state in the time domain.

[0196] In an optional embodiment, the computing module 1130 is further configured to:

[0197] Construct a dynamic model of the steering actuator, which includes the unknown differential torque;

[0198] Construct a sliding surface based on the desired front wheel steering angle and the actual front wheel steering angle;

[0199] By taking the first-order differential of the sliding surface, we obtain the differential equation;

[0200] The differential torque is calculated based on the dynamic model and differential equations.

[0201] In an optional embodiment, the computing module 1130 is further configured to:

[0202] Obtain fault information of the drive actuator and construct the failure mode based on the fault information;

[0203] Construct an optimization objective function based on minimizing tire adhesion utilization according to the failure mode;

[0204] The objective function is solved using a quadratic programming method and a second constraint to obtain the current torque of each wheel.

[0205] In an optional embodiment, the second constraint is created based on the failure mode, total driving torque requirement, differential torque requirement, and the requirement that no additional yaw moment occurs in the effective solution.

[0206] In summary, the actuator fault-tolerant control device for electric vehicles provided in this application can calculate the differential torque based on the actual front wheel angle and the calculated expected front wheel angle when at least one steering actuator fails, causing the vehicle to completely or partially lose its steering ability. When at least one drive actuator fails, the drive torque of each wheel can be reasonably distributed based on the differential torque, thereby enabling the distributed drive electric vehicle to have better tracking performance. The fault-tolerant control method of the hierarchical architecture can effectively improve the fault tolerance and robustness of the steering system, realize reliable control of the vehicle when both the steering actuator and the drive actuator fail simultaneously, and improve driving safety.

[0207] One embodiment of this application provides a computer-readable storage medium storing at least one instruction, which is loaded and executed by a processor to implement the actuator fault-tolerant control method for an electric vehicle as described above.

[0208] One embodiment of this application provides a distributed drive electric vehicle, which includes a processor and a memory. The memory stores at least one instruction, which is loaded and executed by the processor to implement the actuator fault-tolerant control method of the electric vehicle as described above.

[0209] It should be noted that the electric vehicle actuator fault-tolerant control device provided in the above embodiments is only illustrated by the division of the above functional modules when performing actuator fault-tolerant control of electric vehicles. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the electric vehicle actuator fault-tolerant control device can be divided into different functional modules to complete all or part of the functions described above. In addition, the electric vehicle actuator fault-tolerant control device provided in the above embodiments and the electric vehicle actuator fault-tolerant control method embodiments belong to the same concept, and the specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0210] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0211] The above description is not intended to limit the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A fault-tolerant control method for actuators in an electric vehicle, characterized in that, In a distributed drive electric vehicle using front-wheel drive steering, the method includes: Acquire the onboard sensor information and reference trajectory of the distributed drive electric vehicle; The Model Predictive Control (MPC) algorithm is used to track the onboard sensor information and the reference trajectory to obtain the desired front wheel steering angle. When at least one steering actuator malfunction is determined based on the vehicle sensor information, and at least one drive actuator malfunction is determined based on the vehicle sensor information, the sliding mode control (SMC) algorithm is used to calculate the desired front wheel steering angle and the actual front wheel steering angle at the previous moment to obtain the differential torque. The differential torque is calculated based on an optimization allocation algorithm that minimizes tire adhesion utilization, to obtain the current torque of each wheel. The steering of the distributed drive electric vehicle is controlled based on the torque at the current moment; The optimization allocation algorithm based on minimizing tire adhesion utilization calculates the differential torque to obtain the current torque of each wheel, including: acquiring fault information of the drive actuator; constructing failure modes based on the fault information, wherein when the drive actuator partially fails, the fault diagnosis factor in the failure mode takes a value between 0 and 1; constructing an optimization objective function based on minimizing tire adhesion utilization based on the failure mode; solving the optimization objective function using a quadratic programming method and a second constraint to obtain the current torque of each wheel, wherein the second constraint is created based on the failure mode, total driving torque requirement, differential torque requirement, and the requirement that the effective solution does not exhibit additional yaw moment, and the second constraint includes constraints to ensure that the effective solution does not exhibit additional yaw moment: ,in, , , and These are the fault diagnosis factors for each wheel of the vehicle. , , and Let be the torque of each wheel of the vehicle. The wheelbase is the distance between the wheels. The radius is the wheel radius.

2. The actuator fault-tolerant control method for electric vehicles according to claim 1, characterized in that, The step of using the MPC algorithm to track the vehicle sensor information and the reference trajectory to obtain the desired front wheel steering angle includes: Based on the vehicle sensor information and the reference trajectory, an MPC trajectory tracking optimization objective function is constructed. The objective function for optimizing MPC trajectory tracking is transformed into a quadratic function; The quadratic function is solved using a quadratic programming method and the first constraint condition to obtain the optimal control increment sequence in the control time domain; The first n control variables in the optimal control increment sequence are determined as the desired front wheel steering angle, where n is a positive integer.

3. The actuator fault-tolerant control method for electric vehicles according to claim 2, characterized in that, The step of constructing the MPC trajectory tracking optimization objective function based on the vehicle sensor information and the reference trajectory includes: Based on the vehicle sensor information and the reference trajectory, the vehicle dynamics model equation and the trajectory tracking model equation are combined to obtain the state-space expression; Discretize the parameters in the state-space expression to obtain the discretized gain matrix equation; Substituting the discretized gain matrix equation into the state-space expression yields the MPC trajectory tracking optimization objective function.

4. The actuator fault-tolerant control method for electric vehicles according to claim 3, characterized in that, The step of transforming the MPC trajectory tracking optimization objective function into a quadratic function includes: Predict the future state in the time domain based on the MPC trajectory tracking optimization objective function; Generate a quadratic function based on the future state in the time domain.

5. The actuator fault-tolerant control method for electric vehicles according to claim 1, characterized in that, The step of using the SMC algorithm to calculate the differential torque by combining the desired front wheel steering angle and the actual front wheel steering angle at the previous moment includes: Construct a dynamic model of the steering actuator, wherein the dynamic model includes unknown differential torque; Construct a sliding surface based on the desired front wheel steering angle and the actual front wheel steering angle; Taking the first-order differential of the sliding surface, we obtain the differential equation; The differential torque is calculated based on the dynamic model and the differential equation.

6. An actuator fault-tolerant control device for an electric vehicle, characterized in that, For use in a distributed drive electric vehicle employing front-wheel drive steering, the device includes: The acquisition module is used to acquire the on-board sensor information and reference trajectory of the distributed drive electric vehicle; The tracking module is used to track the vehicle sensor information and the reference trajectory using the Model Predictive Control (MPC) algorithm to obtain the desired front wheel steering angle; The calculation module is used to calculate the differential torque by using the sliding mode control (SMC) algorithm to calculate the desired front wheel steering angle and the actual front wheel steering angle at the previous moment when at least one steering actuator failure is determined based on the vehicle sensor information and at least one drive actuator failure is determined based on the vehicle sensor information. The calculation module is also used to calculate the differential torque based on an optimized allocation algorithm that minimizes tire adhesion utilization, so as to obtain the current torque of each wheel. A control module is used to control the steering of the distributed drive electric vehicle based on the torque at the current moment; The calculation module is further configured to: acquire fault information of the drive actuator; construct failure modes based on the fault information, wherein when the drive actuator partially fails, the value of the fault diagnosis factor in the failure mode is between 0 and 1; construct an optimization objective function based on minimizing tire adhesion utilization based on the failure mode; solve the optimization objective function using a quadratic programming method and a second constraint condition to obtain the current torque of each wheel, wherein the second constraint condition is created based on the failure mode, total driving torque requirement, differential torque requirement, and the requirement that the effective solution does not exhibit additional yaw moment, and the second constraint condition includes constraints to ensure that the effective solution does not exhibit additional yaw moment: ,in, , , and These are the fault diagnosis factors for each wheel of the vehicle. , , and Let be the torque of each wheel of the vehicle. The wheelbase is the distance between the wheels. The radius is the wheel radius.

7. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, which is loaded and executed by a processor to implement the actuator fault-tolerant control method for electric vehicles as described in any one of claims 1 to 5.

8. A distributed drive electric vehicle, characterized in that, The distributed drive electric vehicle includes a processor and a memory, wherein the memory stores at least one instruction, which is loaded and executed by the processor to implement the actuator fault-tolerant control method for the electric vehicle as described in any one of claims 1 to 5.

Images