Distributed electric drive vehicle four-wheel differential torque enhanced steering control method and system

Abstract

The invention provides a distributed electric drive vehicle four-wheel differential torque enhanced steering control method and system, and belongs to the technical field of artificial intelligence risk prediction. The vehicle is provided with an upper-layer torque decision module, a lower-layer torque cooperation module and a vehicle-mounted sensor; selecting a pivot steering auxiliary strategy according to an actual driving scene, and pressing a function key to enter a working condition; vehicle dynamic information is obtained through a vehicle-mounted sensor, whether a vehicle is in an in-situ steering state or a small-radius steering state is judged by combining vehicle body static data, and when a selected function key is in an in-situ steering state or a small-radius steering state, the vehicle dynamic information is obtained in real time, and fuzzy reasoning and a neural network algorithm are combined. The ground attachment coefficient and the target yaw angle speed are intelligently judged, and therefore accurate control over the steering torque is achieved; or a double-controller structure is adopted, the torque of each wheel can be effectively distributed and adjusted under complex working conditions, and flexible control and stable running of the vehicle in a narrow space are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of distributed drive vehicle assisted steering, and particularly to a four-wheel differential torque enhanced steering control method and system for a distributed electric drive vehicle. Background Art

[0002] All along, the steering ability of vehicles has been regarded as one of the key indicators for evaluating the performance of new energy vehicles. Among them, the minimum turning radius, as a key parameter for measuring the steering performance of a vehicle, intuitively reflects the flexibility of the vehicle when passing through winding and narrow sections or avoiding impassable obstacles. The smaller the turning radius of a vehicle, the more excellent its maneuverability demonstrated in the steering operation. Most of the new energy passenger vehicles on the market today still use the traditional Ackermann steering system, and its minimum turning radius ranges from approximately 4 to 8 meters, which limits the steering flexibility of the vehicle on complex urban roads.

[0003] To solve this problem, technologies such as in-situ steering and compass steering have emerged, which greatly optimize the steering maneuverability of vehicles. Especially for off-road vehicles performing special tasks, these two technologies endow them with the ability to achieve extremely small radius steering and even quick turning around in streets, bridgeheads, and narrow areas densely covered with obstacles. The above functions mainly rely on complex mechanical steering architectures or independent hydraulic braking systems to achieve wheel locking and steering. In contrast, vehicles driven by in-wheel motors adopt a more concise skid steering principle. Through independent differential steering technology, only by applying appropriate external forces on the left and right wheels can opposite speed differences be generated between them, thereby completing the steering action. This method does not require additional installation of complex steering mechanisms, which not only reduces costs but also significantly improves the flexibility and adaptability of the vehicle.

[0004] In the existing research on distributed drive torque coordination, there is a lack of adaptability of the vehicle to the adhesion conditions of the wild and paved roads. It relies on the driver to control the vehicle state and then adjust the output torque and steering speed, and it is difficult to achieve a good matching effect for the torque coordination of the four in-wheel motors. And for various narrow sections, relying only on traditional steering strategies, in-wheel motor coordination, and driver operations, it is difficult to make the vehicle move quickly, reducing the vehicle's passability for special sections. Therefore, it is very necessary to provide a four-wheel differential torque enhanced steering control method and system for a distributed electric drive vehicle, which adaptively distributes the torque of the four in-wheel motors according to the steering strategy corresponding to the steering demand, optimizes the dynamic response of the torque coordination of each wheel, and ensures the stability and maneuverability of the vehicle state while completing the steering target. Summary of the Invention

[0005] In view of this, the present invention proposes a four-wheel differential torque enhanced steering control method and system for distributed electric drive vehicles, which proposes a wheel hub collaborative control scheme adapted to the ground under limited working conditions with poor outdoor attachment conditions.

[0006] The technical solution of the present invention is realized as follows: On the one hand, the present invention provides a four-wheel differential torque enhanced steering control method for distributed electric drive vehicles. The vehicle is equipped with an upper torque decision-making module, a lower torque coordination module and on-vehicle sensors, and specifically includes the following steps: Judge the steering state through on-vehicle sensors combined with vehicle body static data: If it is in-situ steering, obtain the slip ratio and road surface adhesion coefficient of each wheel, and execute step A; if it is small-radius steering, execute step B; Step A: According to the road surface adhesion coefficient, decide the expected yaw angular velocity that meets the road surface conditions γ desier ; Perform tracking control on the obtained expected yaw angular velocity γ desier to obtain the expected yaw moment M d , and the upper torque decision-making module completes the output of the expected yaw moment M d during steering; Calculate the optimal slip ratio μ - λ according to the curve model of the adhesion coefficient and the slip ratio λ opt ; The lower torque coordination module receives the output request of the expected yaw moment M d , and distributes the output request of the expected yaw moment according to the torque distribution strategy, and outputs the generated result T d to each motor; Step B: Collect the accelerator pedal opening signal and steering wheel angle signal output by the driver through on-vehicle sensors, and decide the longitudinal force to maintain the vehicle's longitudinal driving and the front wheel angle of the vehicle based on the accelerator pedal opening and the steering wheel angle; Calculate the target yaw angular velocity and the target lateral deviation angle of the center of gravity that meet the real-time state of the vehicle; Further obtain the supplementary yaw moment based on the vehicle steering state, establish the constraint conditions based on the expected state objective function to distribute the supplementary moment and set the final torque value of each wheel T i , and request torque output based on the final torque value of each wheel T i , and the motor executes the response to complete small-radius steering.

[0007] On the basis of the above technical solutions, preferably, the steering state is determined by combining in-vehicle sensors with vehicle body static data. Specifically, an inertial navigation system and an MCU calculation unit are also installed on the vehicle. The inertial navigation system, the MCU calculation unit and the in-vehicle sensors obtain and calculate vehicle state parameters: the inertial navigation system obtains the longitudinal and lateral accelerations of the vehicle and the yaw angular velocity, and the in-vehicle sensors obtain the wheel speeds; The MCU calculation unit combines the longitudinal and lateral accelerations of the vehicle, the wheel speeds and the vehicle body static parameters to estimate the real-time slip ratio of each wheel. The slip ratio of each wheel is defined as: ; where λ 1, λ 2, λ 3, λ 4 are the slip ratios of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel respectively; V x , V y are the longitudinal speed and the lateral speed respectively, which are obtained by integrating the longitudinal acceleration and the lateral acceleration of the vehicle respectively; B is the wheelbase of the vehicle; a , b are the distances from the center of mass to the front axle and the rear axle respectively; γ is the yaw angular velocity at the center of mass; ω 1, ω 2, ω 3, ω 4 are the wheel speeds; r is the rolling radius of the tire; Then, the vehicle controller establishes an inference rule corresponding to the slip ratio of each wheel according to the fuzzy experience based on the slip ratio of each wheel of the current vehicle, and estimates the current road surface adhesion coefficient μ i , , where is the adhesion utilization rate of each wheel, is the adhesion confidence of each wheel.

[0008] Preferably, the specific content of the desired yaw angular velocity γ desier that conforms to the road surface conditions is: According to the current road surface adhesion coefficient, the method of fitting the adhesion ellipse is used to define the relationship between the longitudinal force and the lateral force, and combined with the rotational dynamics equation of the wheel, the longitudinal force of the tire is constrained: The yaw angular acceleration at the center of mass based on the adhesion ellipse and the dynamic constraint is defined as: , where , respectively represent the longitudinal force and the lateral force of the left front wheel, the right front wheel, the left rear wheel and the right rear wheel, i = 1, 2, 3, 4; is the yaw moment of inertia of the whole vehicle; is the yaw angular acceleration at the center of mass; The accelerator pedal opening output by the driver collected by the sensor is defined as: ; where is the accelerator pedal opening weight coefficient, and the maximum value is 1; is the accelerator pedal opening signal; is the maximum limit value of the accelerator pedal opening signal; is the free travel of the accelerator pedal opening; Based on the yaw angular acceleration at the center of mass and the accelerator pedal opening, the desired yaw angular velocity that meets the road surface conditions γ desier is defined as: , where represents the unit step response adjustment time under different road surface adhesion limits, is the yaw angular acceleration at the center of mass, is the desired yaw angular velocity γ desier is the derivative of

[0009] Preferably, the specific content of obtaining the desired yaw moment M d is: Based on the obtained desired yaw angular velocity γ desier that meets the road surface conditions, according to the adaptive PID control algorithm BPPID based on the BP neural network, a yaw angular velocity controller with adaptive tuning parameters is established, and the learning rule is defined as follows: ; ; where represents t the adaptive tuning parameter at time represents t the adaptive tuning parameter at time -1, is the learning rate, y is the set value, is the predicted value of the current output, is the feature quantity of the current input; The incremental PID algorithm is defined as: ; ; is the control output at the current time k , is the control output at the previous time k -1, is the increment between the current moment error and the previous moment error; is the current moment k error, and is k the error at the -1 moment and k the error at the -2 moment, 、 and are the proportional gain parameter, integral gain parameter and derivative gain parameter respectively; Based on the above controller tracking control result, the desired target yaw moment M d is obtained and sent to the upper torque decision module.

[0010] Preferably, when there is an error between the actual slip ratio and the desired optimal slip ratio λ opt , the PID control algorithm is adopted, and the desired slip ratio λ opt is used as the reference input, and the wheel drive torque adjustment amount △T is output, and the specific content is: Based on the curve model of the adhesion coefficient and slip ratio under the standard ground, combined with the estimated values of the adhesion coefficients of each wheel obtained above, the optimal slip ratio of the wheel is calculated by performing similarity weight processing on the optimal slip ratio of the standard ground μ - λ λ opt The similarity of the standard ground is defined as: , , is the adhesion coefficient of the current ground, is the adhesion coefficient of the current ground, λ is the average value of the adhesion coefficients of 6 different standard grounds; The optimal slip ratio opt is defined as: where the 、 、 、 、 、 in the numerator are the optimal slip ratios under different standard ground similarities; 、 、 、 、 、 △T are different standard ground similarities; Based on the estimated results of the slip ratios of each wheel obtained above, a PID controller is established, and the optimal slip ratio is used as the reference input to obtain the vehicle driving torque adjustment amount and sent to the upper torque decision module.

[0011] Preferably, the output request for the desired yaw moment is distributed according to a torque distribution strategy, and the result generated by the distribution T d The specific content output to each motor is as follows: The lower-layer torque coordination module receives the target yaw moment request from the upper-layer torque decision module, and the desired yaw moment output by the upper-layer torque decision module M d , that is, the desired target yaw moment output by the adaptive PID control algorithm BPPID based on the BP neural network based on the average distribution rule M d is distributed, and the distribution rule is: , and the distributed torque of each wheel is output to each motor and executed; the lower-layer torque coordination module receives the request for determining the compensation wheel drive torque adjustment amount △T , and at the same time, based on the vehicle real-time state parameters returned by the vehicle-mounted sensor and the vehicle controller, determines the real-time state of the vehicle steering according to the steering stage judgment rule; according to the motor execution torque command situation and the feedback vehicle steering real-time state, as well as the set torque compensation strategy, the wheel drive torque adjustment amount △T intervenes for compensation to maintain the stability of the vehicle state during the in-situ steering process.

[0012] Preferably, in step B, the accelerator pedal opening signal and the steering wheel angle signal output by the driver are collected through the vehicle-mounted sensor, and the longitudinal force for maintaining the longitudinal driving of the vehicle and the front wheel angle of the vehicle are determined based on the accelerator pedal opening and the steering wheel angle. The specific content is that the accelerator pedal opening K 0 and the steering wheel angle θ are collected by the vehicle-mounted sensor, and the maximum total driving torque is formulated based on the external characteristics of the vehicle motor and the limitation of the motor output torque at high and low speeds, where n is the motor speed; based on the linear relationship between the total driving torque and the accelerator pedal, the longitudinal torque that meets the driver's expectation is determined: ; the front wheel angle θ based on the steering wheel angle is calculated and determined through the proportional coefficient between the steering wheel and the front wheel angle .

[0013] Preferably, the calculated target yaw angular velocity and the target lateral deviation angle of the center of gravity that meet the real-time state of the vehicle in step B are specifically as follows: Based on the vehicle controller collecting the vehicle steering state parameters in real time, the MCU calculation unit combines the vehicle real-time state information collected by the vehicle controller with the vehicle static parameters, and the target yaw angular velocity and the target lateral deviation angle of the center of gravity that meet the real-time state of the vehicle are respectively defined as: , , where the stability factor K Defined as , L is the wheelbase, a is the front wheelbase, b is the rear wheelbase, is the average turning angle of the front wheels, is the rear axle wheel cornering stiffness, is the front axle wheel cornering stiffness.

[0014] Preferably, the step B is to obtain a supplementary yaw moment based on the vehicle steering state, establish a constraint condition based on the desired state objective function to distribute the supplementary moment and adjust the final value of each wheel torque. T i Specifically, a dual controller consisting of a fuzzy controller and a sliding mode controller is used to track and adjust the target state parameters. For the fuzzy controller, the fuzzy inference rules and the corresponding membership functions are adjusted based on experimental experience to improve the control output; for the sliding mode controller, a joint control system with yaw angular velocity and lateral deviation angle of the center of gravity as the control targets is designed. The error between the target value and the actual value and the first-order derivative of the error are intermediate variables and are defined as , , where and and are the actual yaw rate and its first-order derivative, and They are the actual lateral deviation angle of the center of gravity and the first-order derivative. The point above the variable represents the first-order derivative. and are the error between the actual yaw rate and the target yaw rate and the first-order derivative of the error respectively; and are the error between the actual lateral deviation angle of the center of gravity and the target lateral deviation angle of the center of gravity and the first-order derivative of the error; considering the robustness requirements of the control, the constant velocity approach rate is selected to design the sliding surface s , ,in k is the weight coefficient, the value range is (0, 1], and it is adjusted according to the actual state of the vehicle; is the relative weight coefficient between the yaw angular velocity error and the inverse, and its value is greater than 0. Finally, the Lyapunov function is defined to test the stability of the joint control system of yaw angular velocity and lateral deviation angle of center of gravity: Based on the tracking control results of the above controller, the vehicle steering and yaw supplementary torques required for the vehicle to complete the steering are obtained respectively. and , is the steering and yaw supplementary torque output by the fuzzy controller, is the additional yaw moment for steering output by the sliding mode controller; based on the obtained additional yaw moment for vehicle steering required for the vehicle to complete steering and , combined with the longitudinal torque that meets the driver's expectation T l , a torque distribution strategy is established as follows: , are the additional yaw moments for steering output by the fuzzy controller and the sliding mode controller respectively, j = 1, 2.

[0015] On the other hand, the present invention also provides a four-wheel differential torque enhanced steering control system for a distributed electric drive vehicle, including an upper-layer torque decision-making module, a lower-layer torque coordination module, and an on-vehicle sensor configured on the vehicle; The on-vehicle sensor is used to obtain vehicle dynamic information after the auxiliary function button is pressed, and combine the vehicle body static data to judge whether the vehicle is in a in-situ steering or small-radius steering state; The upper-layer torque decision-making module, when confirming that the vehicle is in the in-situ steering state, calculates the slip ratio of each wheel, and estimates the current road surface adhesion coefficient based on the fuzzy experience obtained through experiments μ i , calculates the peak yaw angular velocity according to the adhesion ellipse theory, combines the accelerator pedal opening signal output by the driver, and then decides the desired yaw angular velocity that meets the road surface conditions γ desier ; uses the adaptive PID control algorithm BPPID based on the BP neural network, and tunes the proportional, integral, and differential parameters of the PID control by using the neural network information transmission mode and learning rules, and performs tracking control on the obtained desired yaw angular velocity γ desier and obtains the desired yaw moment M d , and the upper-layer torque decision-making module completes the output of the desired yaw moment during steering M d ; calculates the optimal slip ratio according to the μ - λ curve model of the adhesion coefficient and the slip ratio λ opt , when there is an error between the actual slip ratio and the desired optimal slip ratio λ opt , the PID control algorithm is adopted, and the desired slip ratio λ opt is used as the reference input, and the wheel drive torque adjustment amount △T; When it is confirmed that the vehicle is in a small radius turning state, the accelerator pedal opening signal and steering wheel angle signal output by the driver are collected through the on-board sensor, and the longitudinal force and the front wheel angle of the vehicle that maintain the longitudinal travel of the vehicle are decided based on the accelerator pedal opening and the steering wheel angle; the real-time front wheel angle and longitudinal vehicle speed of the vehicle during overload are obtained, and the target yaw rate and target lateral deviation angle of the center of gravity that meet the real-time state of the vehicle are calculated in combination with the static parameters of the vehicle body; based on the calculated values ​​of the target yaw rate and the target lateral deviation angle of the center of gravity, a dual controller consisting of a fuzzy controller constructed based on fuzzy inference rules established based on experimental experience and a synovial controller based on a synovial control algorithm is used to perform weight distribution matching tracking to obtain a supplementary yaw moment based on the steering state of the vehicle; according to the supplementary yaw moment, a constraint condition based on the expected state objective function is established to distribute the supplementary torque and adjust the final value of the torque of each wheel. T i ; The lower torque coordination module receives the desired yaw torque when the vehicle is in a stationary steering state M d The output request of the desired yaw torque is allocated according to the torque allocation strategy, and the result of the allocation is T d Output to each motor, after receiving the compensation wheel drive torque adjustment amount △T After the request is received, the vehicle real-time state parameters returned by the vehicle sensors and the vehicle controller are used to determine the vehicle real-time steering state according to the steering stage judgment rules. The wheel drive torque adjustment amount is calculated based on the motor execution torque command and the feedback vehicle real-time steering state, as well as the set torque compensation strategy. △T Intervene to compensate and maintain the stability of the vehicle during the in-situ steering process; when the vehicle is in a small radius steering state, based on the final value of each wheel torque T i The torque output is requested and the motor executes the response to complete the small radius turn.

[0016] The distributed electric drive vehicle four-wheel differential torque enhancement steering control method and system provided by the present invention have the following beneficial effects compared with the prior art: (1) The present invention proposes a four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle. During the in-situ steering process, by obtaining the vehicle dynamic information in real time and combining fuzzy inference and neural network algorithms, it can intelligently judge the ground adhesion coefficient and the target yaw angular velocity, so as to achieve precise control of the steering torque. In addition, during the small-radius steering control process, a dual controller structure is adopted, which can effectively distribute and adjust the torque of each wheel under complex working conditions, ensuring flexible control and stable driving of the vehicle in a narrow space. Through this innovative control strategy, the present invention significantly improves the power performance and passability of the vehicle under various ground conditions, and enhances the driving safety and comfort. (2) For in-situ steering, on-vehicle sensors obtain vehicle dynamic information such as longitudinal and lateral speeds, longitudinal and lateral accelerations, and the rotational speeds of each wheel, and estimate the current ground adhesion coefficient through fuzzy rules. Secondly, in combination with the accelerator pedal opening, the desired yaw angular velocity is determined through the adhesion ellipse theory, and the driving torque adjustment amount is output with the optimal slip ratio as the control target. Then, an adaptive PID control algorithm based on a BP neural network is used to track the desired yaw angular velocity for the target yaw angular velocity. Finally, based on the steering stage determination of real-time parameters, the yaw moment is distributed and the torque adjustment amount is intervened according to the torque distribution strategy and the torque supplement strategy. The wheel rotational speed decreases and gradually stabilizes, causing the yaw angular velocity to converge to the desired value, thereby improving the consistency during four-wheel steering and maintaining the stability of the vehicle working conditions during in-situ steering. (3) For small-radius steering, the main method is that on-vehicle sensors obtain driver operation signals such as the accelerator pedal opening signal and the steering wheel angle signal, and determine the longitudinal force to maintain vehicle driving and the front wheel angle of the vehicle based on the driver's intention. Secondly, in combination with the vehicle real-time state information obtained by the vehicle controller, such as the front wheel angle, longitudinal vehicle speed, and vehicle static parameters, the target yaw angular velocity and the desired center of gravity lateral deviation angle reflecting the driver's operation are calculated. Then, a dual controller based on a fuzzy controller and a sliding mode controller is used to track and control the target yaw angular velocity and the desired center of gravity lateral deviation angle, and respectively output the additional yaw moment for vehicle steering. Finally, the torque of each wheel is adjusted according to the target torque function of each wheel drive force and the constraint conditions, achieving a significant reduction in the steering radius. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1Schematic diagram of the control strategy for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 2 Schematic diagram of the membership function of the adhesion confidence of each wheel for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 3 Schematic diagram of the membership function of the slip ratio of each wheel for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 4 Schematic diagram of the tracking of the target yaw angular velocity on a high-adhesion ground for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 5 Schematic diagram of the control effect of the wheel speed on a high-adhesion ground for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 6 Schematic diagram of the tracking of the target yaw angular velocity on a low-adhesion ground for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 7 Schematic diagram of the control effect of the wheel speed on a low-adhesion ground for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 8 Architecture diagram of the steering control device for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 9 Fuzzy control system diagram for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 10 Sliding mode control system diagram for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 11 Schematic diagram of the membership function of the error between the actual yaw angular velocity and the target yaw angular velocity for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 12 Schematic diagram of the membership function of the error between the actual lateral deviation angle of the center of gravity and the target lateral deviation angle of the center of gravity for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 13 Schematic diagram of the membership function of the vehicle steering yaw supplementary torque for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 14 High-adhesion ground core displacement for a four-wheel differential torque enhanced steering control method and system of a distributed electric drive vehicle according to the present invention; Figure 15 This is the low-adhesion ground core displacement of a four-wheel differential torque enhanced steering control method and system for a distributed electric drive vehicle according to the present invention. Specific embodiments

[0019] Next, in combination with the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] In the current research on distributed electric drive torque coordination, there is a lack of adaptability to the adhesion conditions of the wild and paved ground, and it relies on the driver's judgment of the vehicle state to adjust the output torque and steering speed, resulting in poor torque coordination effect of the four-wheel hub motors. In view of this, as Figure 1 shown, on the one hand, the present invention provides a four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle. The vehicle is equipped with an upper torque decision-making module, a lower torque coordination module, and on-vehicle sensors, and specifically includes the following steps: S1: Select the in-situ steering assistance strategy according to the actual driving scenario, and press the function button to enter the working condition; S2: Obtain the vehicle dynamic information through the on-vehicle sensors, and combine the vehicle static data to judge whether the vehicle is in the in-situ steering or small-radius steering state. When the selected function button is in-situ steering, calculate the slip ratio of each wheel, and estimate the current road surface adhesion coefficient based on the fuzzy experience obtained from experiments to establish the inference rules corresponding to the slip ratio of each wheel μ i , and further execute step A; when the selected function button is in the small-radius steering state, execute step B.

[0021] In this step, the vehicle dynamic information is obtained through the on-vehicle sensors, and the vehicle is judged to be in the in-situ steering or small-radius steering state by combining the vehicle static data. Specifically, an inertial navigation system, an MCU calculation unit, and a vehicle controller are also installed on the vehicle. The inertial navigation system, the MCU calculation unit, and the on-vehicle sensors obtain and calculate the vehicle state parameters: the inertial navigation system obtains the vehicle longitudinal and lateral accelerations and the yaw angular velocity, and the on-vehicle sensors obtain the wheel speeds; The MCU calculation unit combines the vehicle longitudinal and lateral accelerations, the wheel speeds, and the vehicle static parameters to estimate the real-time slip ratio of each wheel. The slip ratio of each wheel is defined as:

[0022] Among them, λ 1. λ 2. λ 3.λ are the slip ratios of the left front wheel, right front wheel, left rear wheel, and right rear wheel, respectively; V x , V y are the longitudinal speed and lateral speed, respectively, obtained by integrating the vehicle's longitudinal acceleration and lateral acceleration; B is the vehicle track width; a , b are the distances from the center of mass to the front axle and rear axle, respectively; γ is the yaw rate at the center of mass; ω 1, ω 2, ω 3, ω 4 are the wheel speeds; r is the tire rolling radius; Then, the vehicle controller establishes inference rules for the slip ratios of each wheel of the current vehicle according to fuzzy experience, and estimates the current road surface adhesion coefficient μ i , , where is the adhesion utilization rate of each wheel, is the adhesion confidence of each wheel, i = 1, 2, 3, 4; among them, inference rules for the adhesion confidence of each wheel and the slip ratio of each wheel are established according to fuzzy experience, referring to Appendix Figure 2 and Appendix Figure 3 as shown.

[0023] Step A can be specifically divided into: A100: According to the obtained road surface adhesion coefficient μ i , calculate the peak yaw rate according to the adhesion ellipse theory, and combine with the accelerator pedal opening signal output by the driver, and then decide the expected yaw rate γ desier that meets the road surface conditions.

[0024] In step A100, the specific content of deciding the expected yaw rate γ desier that meets the road surface conditions is: According to the current road surface adhesion coefficient, use the method of adhesion ellipse fitting to define the relationship between the longitudinal force and the lateral force, and combine with the rotational dynamics equation of the wheel to constrain the tire longitudinal force: The yaw angular acceleration at the center of mass based on the adhesion ellipse and dynamic constraints is defined as: , where , respectively represent the longitudinal force and lateral force of the left front wheel, right front wheel, left rear wheel, and right rear wheel, i = 1, 2, 3, 4; is the yaw moment of inertia of the whole vehicle; is the yaw angular acceleration at the center of mass.

[0025] The accelerator pedal opening output by the driver collected by the sensor is defined as: ; where is the accelerator pedal opening weight coefficient, and the maximum value is 1; is the accelerator pedal opening signal; is the maximum limit value of the accelerator pedal opening signal; is the free travel of the accelerator pedal; Based on the yaw angular acceleration and accelerator pedal opening at the center of mass, the expected yaw angular velocity that meets the road surface conditions γ desier is defined as: , where represents the unit step response adjustment time under different road surface adhesion limits, is the yaw angular acceleration at the center of mass, is the expected yaw angular velocity γ desier is the derivative of.

[0026] A200: Use the adaptive PID control algorithm BPPID based on the BP neural network, and use the neural network information transfer mode and learning rules to tune the proportional, integral, and differential parameters of the PID control, and for the obtained expected yaw angular velocity γ desier perform tracking control and obtain the expected yaw moment M d , and the upper torque decision module completes the output of the expected yaw moment M d during steering.

[0027] In step A200, the specific content of obtaining the expected yaw moment M d is: Based on the obtained expected yaw angular velocity that meets the road surface conditions γ desier , according to the adaptive PID control algorithm BPPID based on the BP neural network, establish a yaw angular velocity controller with adaptive tuning parameters, and the learning rules are defined as follows: ; ; where represents t the adaptive tuning parameter at time, represents t the adaptive tuning parameter at -1 time, is the learning rate, y is the set value, is the predicted value of the current output, is the feature quantity of the current input; The incremental PID algorithm is defined as: ; ; is the current moment k of the control output, is the control output of the previous moment k -1 of the control output, is the increment between the error at the current moment and the error at the previous moment; is the current moment k of the error, and is k -1 moment and k -2 moment of the error, 、 and are the proportional gain parameter, integral gain parameter and derivative gain parameter respectively; Based on the above controller tracking control results, the desired target yaw moment M d is obtained and sent to the upper torque decision module.

[0028] A300: According to the μ - λ curve model of the adhesion coefficient and slip ratio, calculate the optimal slip ratio λ opt , when there is an error between the actual slip ratio and the desired optimal slip ratio λ opt , adopt the PID control algorithm, with the desired slip ratio λ opt as the reference input, and output the wheel drive torque adjustment amount △T .

[0029] In step A300, the specific content of the wheel drive torque adjustment amount △T is: Based on the μ - λ curve model of the adhesion coefficient and slip ratio under the standard ground, combined with the estimated values of the adhesion coefficients of each wheel obtained above, calculate the wheel optimal slip ratio λ opt by performing similarity weight processing on the optimal slip ratio of the standard ground, and define the standard ground similarity as: , , is the adhesion coefficient of the current ground, is the adhesion coefficient of the current ground, is the mean of the adhesion coefficients of 6 different standard grounds; the optimal slip ratio λ opt is defined as: , where in the numerator , , , , , are the optimal slip ratios under different standard ground similarities; , , , , , are different standard ground similarities; Based on the estimated results of each wheel slip ratio obtained above, a PID controller is established, and the optimal slip ratio is used as the reference input to obtain the vehicle driving torque adjustment △T , and send it to the upper torque decision module.

[0030] A400: The lower torque coordination module receives the output request of the desired yaw moment M d , and distributes the output request of the desired yaw moment according to the torque distribution strategy, and outputs the generated result T d to each motor.

[0031] The specific content of step A400 is: The lower torque coordination module receives the target yaw moment request from the upper torque decision module, and the desired yaw moment M d output by the upper torque decision module, that is, the desired target yaw moment M d output by the adaptive PID control algorithm BPPID based on the BP neural network is distributed according to the average distribution rule. The distribution rule is: , and the distributed torque of each wheel is output to each motor and executed.

[0032] A500: The lower torque coordination module receives the request to determine the compensation wheel drive torque adjustment Δ T , and at the same time, based on the vehicle real-time state parameters returned by the vehicle-mounted sensor and the vehicle controller, determines the real-time steering state of the vehicle according to the steering stage judgment rule.

[0033] Based on the vehicle state obtained by the vehicle-mounted sensor and the estimation result of the real-time state by the vehicle controller, the vehicle state is divided into a starting stage, an elastic deformation stage, a side-slip stage, and a steady-state yaw stage according to the steering determination rule, providing a basis for formulating subsequent supplementary strategies.

[0034] A600: Based on the motor's execution of the torque command, the real-time state of vehicle steering feedback, and the set torque compensation strategy, the wheel drive torque adjustment amount △T intervenes for compensation to maintain the stability of the vehicle state during the in-situ steering process.

[0035] Monitor the real-time state parameters during the vehicle steering process based on in-vehicle sensors and the vehicle controller. Determine the intervention timing of the wheel drive torque adjustment amount △T based on the established drive compensation strategy to improve the consistency during four-wheel steering and maintain the working condition stability. Based on the above-established control structure, conduct experimental verification on high- and low-adhesion grounds. The experimental results are as Figure 4 and Figure 5 shown. On high-adhesion ground, as the vehicle gradually starts, the yaw angular velocity approaches the expected value. At 5 to 6 s, the yaw angular velocity overshoots. At this time, the supplementary control intervenes, the wheel speed drops and gradually stabilizes, causing the yaw angular velocity to converge to the expected value of 8° / s. On low-adhesion ground, the vehicle's starting response is more rapid and gradually approaches the expected yaw angular velocity value. A small overshoot occurs at 5 to 6 s. After the supplementary control intervenes, the yaw angular velocity gradually converges to the expected value.

[0036] Please refer to the appendix Figure 6 , for the restricted working conditions with a narrow steering space, the present invention proposes a flexible steering control scheme with supplementary enhanced steering, that is, small-radius steering.

[0037] Step B can be specifically divided into: B100: Collect the accelerator pedal opening signal and steering wheel angle signal output by the driver through in-vehicle sensors, and determine the longitudinal force to maintain the vehicle's longitudinal driving and the front-wheel angle of the vehicle based on the accelerator pedal opening and steering wheel angle.

[0038] Specifically, use in-vehicle sensors to collect the accelerator pedal opening K 0 and the steering wheel angle θ , and formulate the maximum total driving torque based on the external characteristics of the vehicle motor and the limitation of the motor output torque at high and low speeds, where n is the motor speed; based on the linear relationship between the total driving torque and the accelerator pedal, determine the longitudinal torque that meets the driver's expectation: ; calculate and determine the front-wheel angle θ based on the steering wheel angle through the proportional coefficient between the steering wheel and the front-wheel angle.

[0039] B200: Obtain the real-time front-wheel angle and longitudinal vehicle speed during vehicle driving overload, and combine the static parameters of the vehicle body, such as wheelbase, track width, vehicle mass, tire cornering stiffness, and rolling radius, to calculate the target yaw angular velocity and target center-of-gravity lateral deviation angle that conform to the real-time state of the vehicle.

[0040] Specifically, based on the vehicle steering state parameters collected in real time by the vehicle control unit, the MCU calculation unit combines the real-time state information of the vehicle collected by the vehicle control unit with the static parameters of the vehicle to obtain the target yaw angular velocity and the target centroid lateral deviation angle that conform to the real-time state of the vehicle which are respectively defined as: , , where the stability factor K is defined as , L is the wheelbase, a is the front wheelbase, b is the rear wheelbase, is the average front wheel angle, is the cornering stiffness of the rear axle wheels, is the cornering stiffness of the front axle wheels. A dual controller composed of a fuzzy controller and a sliding mode controller is used to track and tune the target state parameters. The two control structures are shown in Figure 7 and 8 respectively.

[0041] B300: Based on the calculated values of the target yaw angular velocity and the target centroid lateral deviation angle, a dual controller composed of a fuzzy controller constructed by using fuzzy inference rules established based on experimental experience and a sliding mode controller based on the sliding mode control algorithm is used for weight distribution matching tracking to obtain the supplementary yaw moment based on the vehicle steering state. For the fuzzy controller, the fuzzy inference rules and the corresponding membership functions are tuned based on experimental experience as shown in Figure 9 and Figure 10 and Figure 11 respectively.

[0042] Specifically, a dual controller composed of a fuzzy controller and a sliding mode controller is used to track and tune the target state parameters. Among them, for the fuzzy controller, the fuzzy inference rules and the corresponding membership functions are tuned based on experimental experience to improve the control output; for the sliding mode controller, a combined control system with the yaw angular velocity and the centroid lateral deviation angle as the control targets is designed. The errors between the target values and the actual values of the two and the first derivative of the error are defined as intermediate variables , , where and are the actual yaw angular velocity and its first derivative respectively, and are the actual centroid lateral deviation angle and its first derivative respectively. The dot above the variable represents the first derivative, and are the error between the actual yaw angular velocity and the target yaw angular velocity and the first derivative of the error respectively; and ​They are the error between the actual lateral deviation angle of the center of gravity and the target lateral deviation angle of the center of gravity, and the first derivative of the error. Considering the requirements of control robustness, a constant velocity reaching law is selected to design the sliding surface. s , , where k is the weight coefficient, and its value range is (0, 1], which is adjusted according to the actual state of the vehicle. is the relative weight coefficient between the yaw rate error and its derivative, and its value is greater than 0. Finally, the Lyapunov function is defined to verify the stability of the joint control system of the yaw rate and the lateral deviation angle of the center of gravity: ; Based on the tracking control results of the above controller, the additional yaw moment required for the vehicle to complete the steering and are obtained respectively. is the additional yaw moment output by the fuzzy controller, is the additional yaw moment output by the sliding mode controller.

[0043] B400: According to the additional yaw moment, establish the constraint conditions based on the expected state objective function to distribute the additional moment and adjust the final value of each wheel torque T i .

[0044] Specifically, based on the additional yaw moment and required for the vehicle to complete the steering, combined with the longitudinal moment T l that meets the driver's expectation, a torque distribution strategy is established as follows: , are the additional yaw moments output by the fuzzy controller and the sliding mode controller respectively, j = 1, 2.

[0045] Based on the above distribution rules and the dual controllers, experiments are verified on high-friction ground and low-friction ground respectively. The experimental results are as Figure 12 and Figure 13 shown. Under high-friction conditions, compared with the steering condition without additional intervention, when the outer steering wheel is supplemented and enhanced, the turning radius is effectively reduced by at least more than 20%. Under low-friction conditions, when the outer steering wheel is supplemented and enhanced, the turning radius is effectively reduced by at least more than 15%, and the tracking error is effectively reduced.

[0046] B500: Based on the final value of each wheel torque T i request torque output, and the motor executes the response to complete the small-radius steering.

[0047] In step B of this control method, first, in-vehicle sensors obtain driver operation signals such as the accelerator pedal opening signal and the steering wheel angle signal, and based on the driver's intention, the longitudinal force to maintain vehicle driving and the front wheel angle of the vehicle are determined; secondly, combined with the vehicle real-time state information obtained by the vehicle controller such as the front wheel angle, longitudinal vehicle speed, and vehicle static parameters, the target yaw angular velocity and the expected center of gravity lateral deviation angle reflecting the driver's operation are calculated; then, a dual controller based on a fuzzy controller and a sliding mode controller is used to perform tracking control on the target yaw angular velocity and the expected center of gravity lateral deviation angle and output the vehicle steering supplementary yaw moment; finally, the torque of each wheel is adjusted according to the target torque function of each wheel drive force and the constraint conditions. By adopting the dual controller structure, it is possible to effectively distribute and adjust the torque of each wheel under complex working conditions, ensuring the flexible control and stable driving of the vehicle in a narrow space. Through this innovative control strategy, the present invention significantly improves the power performance and passability of the vehicle under various ground conditions, and enhances the driving safety and comfort.

[0048] In addition, the present invention also provides a four-wheel differential torque enhanced steering control system for a distributed electric drive vehicle, including an upper torque decision-making module, a lower torque coordination module, and in-vehicle sensors configured on the vehicle; The in-vehicle sensors are used to obtain vehicle dynamic information after the auxiliary function button is pressed, and combined with the vehicle body static data to determine whether the vehicle is in a state of in-situ steering or small-radius steering; When the upper torque decision-making module confirms that the vehicle is in the in-situ steering state, it calculates the slip ratio of each wheel, and based on the fuzzy experience obtained through experiments, establishes an inference rule corresponding to the slip ratio of each wheel to estimate the current road surface adhesion coefficient μ i , calculates the peak yaw angular velocity according to the adhesion ellipse theory, and combined with the accelerator pedal opening signal output by the driver, further determines the expected yaw angular velocity that conforms to the road surface conditions γ desier ; uses the adaptive PID control algorithm BPPID based on the BP neural network, and tunes the three parameters of proportional, integral, and differential of the PID control by using the neural network information transmission mode and learning rules for the obtained expected yaw angular velocity γ desier for tracking control and obtains the expected yaw moment M d , and the upper torque decision-making module completes the output of the expected yaw moment during steering M d ; calculates the optimal slip ratio according to the μ - λ curve model of the adhesion coefficient and the slip ratio λ opt , when the actual slip ratio and the expected optimal slip ratio λopt When there is an error, the PID control algorithm is used to control the desired slip rate. λ opt As a reference input, the output wheel drive torque adjustment △T ; When it is confirmed that the vehicle is in a small radius turning state, the accelerator pedal opening signal and steering wheel angle signal output by the driver are collected through the on-board sensor, and the longitudinal force and the front wheel angle of the vehicle that maintain the longitudinal travel of the vehicle are decided based on the accelerator pedal opening and the steering wheel angle; the real-time front wheel angle and longitudinal vehicle speed of the vehicle during overload are obtained, and the target yaw rate and target lateral deviation angle of the center of gravity that meet the real-time state of the vehicle are calculated in combination with the static parameters of the vehicle body; based on the calculated values ​​of the target yaw rate and the target lateral deviation angle of the center of gravity, a dual controller consisting of a fuzzy controller constructed based on fuzzy inference rules established based on experimental experience and a synovial controller based on a synovial control algorithm is used to perform weight distribution matching tracking to obtain a supplementary yaw moment based on the steering state of the vehicle; according to the supplementary yaw moment, a constraint condition based on the expected state objective function is established to distribute the supplementary torque and adjust the final value of the torque of each wheel. T i ; The lower torque coordination module receives the desired yaw torque when the vehicle is in a stationary steering state M d The output request of the desired yaw torque is allocated according to the torque allocation strategy, and the result of the allocation is T d Output to each motor, after receiving the compensation wheel drive torque adjustment amount △T After the request is received, the vehicle real-time state parameters returned by the vehicle sensors and the vehicle controller are used to determine the vehicle real-time steering state according to the steering stage judgment rules. The wheel drive torque adjustment amount is calculated based on the motor execution torque command and the feedback vehicle real-time steering state, as well as the set torque compensation strategy. △T Intervene to compensate and maintain the stability of the vehicle during the in-situ steering process; when the vehicle is in a small radius steering state, based on the final value of each wheel torque T i The torque output is requested and the motor executes the response to complete the small radius turn.

[0049] The above description is only a preferred embodiment 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 protection scope of the present invention.

Claims

1. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle, characterized in that, The vehicle is equipped with an upper-layer torque decision-making module, a lower-layer torque coordination module, and on-vehicle sensors. The specific steps are as follows: Judge the steering state by combining the on-vehicle sensors with the vehicle body static data: If it is a stationary turn, obtain the slip ratio of each wheel and the road surface adhesion coefficient, and execute step A; if it is a small-radius turn, execute step B; Step A: Determine the expected yaw rate that meets the road surface conditions based on the road surface adhesion coefficient γ desier ; Perform tracking control on the obtained expected yaw rate γ desier to obtain the expected yaw moment M d , and the output of the expected yaw moment M d during steering is completed by the upper torque decision module; Calculate the optimal slip ratio μ - λ based on the curve model of the adhesion coefficient and the slip ratio λ opt ; The lower-layer torque coordination module receives the output request of the desired yaw moment M d and distributes the output request of the desired yaw moment according to the torque distribution strategy, and outputs the generated result T d to each motor; Step B: Collect the accelerator pedal opening signal and the steering wheel angle signal output by the driver through in-vehicle sensors, and determine the longitudinal force for maintaining the longitudinal driving of the vehicle and the front wheel angle of the vehicle based on the accelerator pedal opening and the steering wheel angle; calculate the target yaw angular velocity and the target lateral center of gravity deviation angle that conform to the real-time state of the vehicle; further obtain the supplementary yaw moment based on the steering state of the vehicle, establish the constraint conditions based on the desired state objective function to distribute the supplementary moment and set the final torque value of each wheel. T i , based on the final torque value of each wheel T i Request torque output, and the motor executes the response to complete the small-radius steering.

2. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 1, characterized in that Judge the steering state by combining the on-vehicle sensors with the vehicle body static data. Specifically, an inertial navigation system and an MCU calculation unit are also installed on the vehicle. The inertial navigation system, the MCU calculation unit, and the on-vehicle sensors obtain and calculate the vehicle state parameters: The inertial navigation system obtains the longitudinal and lateral accelerations of the vehicle and the yaw angular velocity, and the on-vehicle sensors obtain the wheel speeds; The MCU calculation unit combines the longitudinal and lateral accelerations of the vehicle, the wheel speeds, and the static body parameters to estimate the real-time slip ratio of each wheel. The slip ratio of each wheel is defined as: ; where λ 1、 λ 2、 λ 3、 λ 4 are the slip ratios of the left front wheel, right front wheel, left rear wheel, and right rear wheel respectively; V x 、 V y are the longitudinal speed and lateral speed respectively, obtained by integrating the longitudinal acceleration and lateral acceleration of the vehicle; B is the wheelbase of the vehicle; a 、 b are the distances from the center of mass to the front axle and rear axle respectively; γ is the yaw angular velocity at the center of mass; ω 1、 ω 2、 ω 3、 ω 4 are the wheel speeds; r is the tire rolling radius; Then, according to the slip ratios of the wheels of the current vehicle, the vehicle controller establishes inference rules corresponding to the slip ratios of each wheel based on fuzzy experience and estimates the current road surface adhesion coefficient μ i , , where is the adhesion utilization rate of each wheel, is the adhesion confidence of each wheel.

3. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 2, wherein Determine the expected yaw rate that conforms to the road surface conditions γ desier The specific content is as follows: According to the current road surface adhesion coefficient, use the method of adhesion ellipse fitting to define the relationship between the longitudinal force and the lateral force, and combine the rotational dynamics equation of the wheel to constrain the tire longitudinal force: The yaw angular acceleration at the centroid based on the adhesion ellipse and the dynamic constraints is defined as: , where and represent the longitudinal and lateral forces of the left front wheel, right front wheel, left rear wheel and right rear wheel respectively, i = 1, 2, 3, 4; is the yaw moment of inertia of the whole vehicle; is the yaw angular acceleration at the center of mass; The accelerator pedal opening output by the driver collected by the sensor is defined as: ; where is the acceleration pedal opening weight coefficient, and the maximum value is 1; is the acceleration pedal opening signal; is the maximum limit value of the acceleration pedal opening signal; is the free travel of the acceleration pedal; Based on the yaw angular acceleration at the center of mass and the acceleration pedal opening, the desired yaw angular velocity that meets the road surface conditions γ desier is defined as: , where represents the unit step response adjustment time under different road surface adhesion limits, is the yaw angular acceleration at the center of mass, is the desired yaw angular velocity γ desier the derivative of.

4. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 3, characterized in that, Obtain the desired yaw moment M d The specific content is as follows: Based on the obtained desired yaw rate that conforms to the road surface conditions γ desier , according to the adaptive PID control algorithm BPPID based on the BP neural network, a yaw rate controller with self-tuning parameters is established, and the learning rules are defined as follows: ; ; where represents t the adaptive tuning parameter at time represents t the adaptive tuning parameter at time - 1, is the learning rate, y is the set value, is the predicted value of the current output, is the feature quantity of the current input; The incremental PID algorithm is defined as: ； ； is the control output at the current moment k ; is the control output at the previous moment k minus 1; is the increment between the error at the current moment and the error at the previous moment; is the error at the current moment k ; and are k the errors at the moment minus 1 and k the moment minus 2; , and are the proportional gain parameter, the integral gain parameter, and the derivative gain parameter respectively; Based on the above controller tracking control results, the desired target yaw moment M d is obtained and sent to the upper torque decision module.

5. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 4, characterized in that When the actual slip ratio and the desired optimal slip ratio λ opt have an error, the PID control algorithm is adopted, and the desired slip ratio λ opt is used as the reference input to output the adjustment amount of the wheel drive torque △T , and the specific content is as follows: Based on the adhesion coefficient and slip ratio under standard ground μ - λ Curve model, combined with the estimated values of adhesion coefficients of each wheel obtained, calculates the optimal slip ratio of the wheel by performing similarity weight processing on the optimal slip ratio of the standard ground λ opt , defines the similarity of the standard ground as: , , is the adhesion coefficient of the current ground, is the adhesion coefficient of the current ground, is the average value of the adhesion coefficients of 6 different standard grounds; defines the optimal slip ratio λ opt as: , where in the numerator , , , , , are the optimal slip ratios under different standard ground similarities; , , , , , are different standard ground similarities; based on the estimated results of the slip ratios of each wheel obtained, a PID controller is established, and the optimal slip ratio is used as the reference input to obtain the vehicle driving torque adjustment amount △T , and send it to the upper torque decision-making module.

6. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 5, characterized in that, The output request for the desired yaw moment is distributed according to the torque distribution strategy, and the result generated by the distribution T d The specific content output to each motor is as follows: The lower-layer torque coordination module receives the target yaw moment request from the upper-layer torque decision module, and the desired yaw moment output by the upper-layer torque decision module M d , that is, the desired target yaw moment output by the adaptive PID control algorithm BPPID based on the BP neural network based on the average distribution rule M d is distributed, and the distribution rule is: , and the torque distributed to each wheel is output to each motor and executed; The lower-layer torque coordination module receives a request to determine the adjustment amount of the compensated wheel drive torque, and at the same time, based on the real-time vehicle state parameters returned by in-vehicle sensors and the vehicle controller, determines the real-time steering state of the vehicle according to the steering stage judgment rule; according to the motor execution torque command situation, the real-time steering state of the feedback vehicle, and the set torque compensation strategy, the adjustment amount of the wheel drive torque △T intervenes for compensation to maintain the stability of the vehicle state during the in-situ steering process. △T ​ 7. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 2, characterized in that, In step B, the accelerator pedal opening signal and the steering wheel angle signal output by the driver are collected by in-vehicle sensors, and the longitudinal force for maintaining the longitudinal driving of the vehicle and the front wheel angle of the vehicle are determined based on the accelerator pedal opening and the steering wheel angle. Specifically, the in-vehicle sensors are used to collect the accelerator pedal opening K 0 and the steering wheel angle θ , and the maximum total driving torque is formulated based on the external characteristics of the vehicle motor and the limitation of the motor output torque at high and low speeds , where n is the motor speed; Based on the linear relationship between the total driving torque and the accelerator pedal, make a decision on the longitudinal torque that meets the driver's expectations : ; Calculate and determine the front wheel angle based on the steering wheel angle through the proportionality coefficient between the steering wheel and the front wheel angle θ of the front wheel angle .

8. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 7, characterized in that, The calculated target yaw angular velocity and target lateral center-of-gravity deviation angle described in step B are specifically as follows. Based on the vehicle steering state parameters collected in real time by the vehicle control unit, the MCU calculation unit combines the vehicle static parameters with the vehicle real-time state information collected by the vehicle control unit to obtain the target yaw angular velocity and the target lateral center-of-gravity deviation angle which are respectively defined as: , where the stability factor K is defined as , L is the wheelbase, a is the front wheelbase, b is the rear wheelbase, is the average front wheel angle, is the cornering stiffness of the rear axle wheels, is the cornering stiffness of the front axle wheels.

9. A four-wheel differential torque enhanced steering control method for a distributed electric drive vehicle according to claim 8, characterized in that Obtaining the supplementary yaw moment based on the vehicle steering state in step B, establishing the constraint conditions based on the desired state objective function to distribute the supplementary moment and determine the final torque value of each wheel T i , specifically, a dual controller composed of a fuzzy controller and a sliding mode controller is adopted to track and tune the target state parameters. Among them, for the fuzzy controller, the fuzzy inference rules and the corresponding membership functions are tuned based on experimental experience to improve the control output; for the sliding mode controller, a combined control system with the yaw angular velocity and the lateral deviation angle of the center of gravity as the control objectives is designed, and the error between the target value and the actual value of the two and the first derivative of this error are defined as intermediate variables as , , where and and are the actual yaw angular velocity and its first derivative respectively, and are the actual lateral deviation angle of the center of gravity and its first derivative respectively, and the dot above the variable represents the first derivative, and are the error between the actual yaw angular velocity and the target yaw angular velocity and the first derivative of the error respectively; and are the error between the actual lateral deviation angle of the center of gravity and the target lateral deviation angle of the center of gravity and the first derivative of the error respectively; Considering the robustness requirements of the control, a constant velocity reaching rate is selected to design the sliding mode surface s , , where k is the weight coefficient, and its value range is (0, 1], which is tuned according to the actual state of the vehicle; is the relative weight coefficient between the yaw angular velocity error and its reciprocal, and its value is greater than 0; Finally, the Lyapunov function is defined to verify the stability of the combined control system of the yaw angular velocity and the lateral deviation angle of the center of gravity: ; Based on the tracking control results of the above controller, the vehicle steering yaw supplementary moments and required for the vehicle to complete the steering are obtained respectively, is the steering yaw supplementary moment output by the fuzzy controller, is the steering yaw supplementary moment output by the sliding mode controller; Based on the obtained vehicle steering yaw supplementary moments and , combined with the longitudinal torque T l that meets the driver's expectation, a torque distribution strategy is established as follows: , are the steering yaw supplementary moments output by the fuzzy controller and the sliding mode controller respectively, j = 1, 2.

10. A four-wheel differential torque enhanced steering control system for a distributed electric drive vehicle, characterized in that, It includes an upper-layer torque decision-making module, a lower-layer torque coordination module, and on-vehicle sensors configured on the vehicle; The on-vehicle sensors are used to obtain the vehicle dynamic information after the auxiliary function button is pressed, and judge whether the vehicle is in a stationary turn or a small-radius turn state by combining the vehicle body static data; When the upper torque decision-making module confirms that the vehicle is in the in-situ steering state, it calculates the slip ratio of each wheel, and estimates the current road surface adhesion coefficient based on the fuzzy experience obtained from experiments, μ i calculates the peak yaw angular velocity according to the adhesion ellipse theory, and combines with the accelerator pedal opening signal output by the driver to further decide the expected yaw angular velocity that meets the road conditions. γ desier Using the adaptive PID control algorithm BPPID based on the BP neural network, the proportional, integral, and differential parameters of the PID control are tuned by using the neural network information transfer mode and learning rules, and the obtained expected yaw angular velocity γ desier is tracked and controlled to obtain the expected yaw moment. M d The upper torque decision-making module completes the output of the expected yaw moment during steering. M d According to the curve model of the adhesion coefficient and the slip ratio μ - λ the optimal slip ratio is calculated. λ opt When there is an error between the actual slip ratio and the expected optimal slip ratio λ opt the PID control algorithm is adopted, and the expected slip ratio λ opt is used as the reference input to output the wheel drive torque adjustment amount. △T When the vehicle is confirmed to be in the small-radius steering state, the accelerator pedal opening signal and the steering wheel angle signal output by the driver are collected by in-vehicle sensors, and the longitudinal force to maintain the vehicle's longitudinal driving and the front wheel angle of the vehicle are decided based on the accelerator pedal opening and the steering wheel angle; the real-time front wheel angle and longitudinal vehicle speed during the vehicle's driving overload are obtained, and combined with the body static parameters, the target yaw angular velocity and the target center-of-gravity lateral deviation angle that meet the real-time state of the vehicle are calculated. Based on the calculated values of the target yaw angular velocity and the target lateral deviation angle of the center of gravity, a dual controller composed of a fuzzy controller constructed by using fuzzy inference rules established based on experimental experience and a sliding mode controller based on the sliding mode control algorithm is used for weight allocation matching tracking to obtain a supplementary yaw moment based on the vehicle steering state; according to the supplementary yaw moment, constraint conditions based on the desired state objective function are established to allocate the supplementary moment and tune the final torque value of each wheel T i ; The lower torque coordination module receives the desired yaw torque when the vehicle is in a stationary steering state M d The output request of the desired yaw torque is allocated according to the torque allocation strategy, and the result of the allocation is T d Output to each motor, after receiving the compensation wheel drive torque adjustment amount △T After the request is received, the vehicle real-time state parameters returned by the vehicle sensors and the vehicle controller are used to determine the vehicle real-time steering state according to the steering stage judgment rules. The wheel drive torque adjustment amount is calculated based on the motor execution torque command and the feedback vehicle real-time steering state, as well as the set torque compensation strategy. △T Intervene to compensate and maintain the stability of the vehicle during the in-situ steering process; when the vehicle is in a small radius steering state, based on the final value of each wheel torque T i The torque output is requested and the motor executes the response to complete the small radius turn.

Images