Torque distribution control method for four-wheel hub motor driven intelligent vehicle

Abstract

This invention discloses a torque distribution control method for a four-wheel hub motor driven intelligent vehicle, relating to the field of pure electric vehicle control. The invention selects two characteristic quantities: the vehicle's center of gravity sideslip angle and its angular velocity. Utilizing a two-degree-of-freedom nonlinear dynamics model of the vehicle, initial values ​​of the vehicle's center of gravity sideslip angle and yaw rate are input to construct a phase plane of the vehicle's center of gravity sideslip angle and sideslip rate. The stable driving region and extension set of the vehicle are determined. Based on extension control theory, the dynamic weighting coefficients of the objective function for hub motor torque distribution control based on stability and energy saving are determined. The hub motor torque is dynamically distributed in real time, taking into account both the vehicle's driving stability and economy. When using this method, the reasonable distribution of torque among the hub motors in the four-wheel hub motor driven intelligent vehicle not only reduces energy loss but also ensures stable vehicle operation.

Description

Technical Field

[0001] This invention belongs to the field of pure electric vehicle control, specifically relating to a torque distribution control method for a four-wheel hub motor driven intelligent vehicle. Background Technology

[0002] Many scholars employ a hierarchical control strategy for the stability control of four-wheel hub electric drive vehicles. The upper-level control primarily aims to obtain the generalized target force or torque for stable vehicle operation. Since the four wheels of a four-wheel hub electric drive vehicle are independent, torque distribution to achieve the upper-level control strategy falls under the lower-level control. Currently, the main distribution methods are average distribution, dynamic load distribution, and optimal distribution.

[0003] Jin L et al. from Jilin University evenly distributed the total longitudinal torque obtained from longitudinal speed PID control to each wheel, and then distributed the desired yaw torque from the upper layer to the left and right wheels according to the wheel steering angle, ensuring the power requirements for longitudinal drivability and lateral handling of the vehicle. Aria N et al. from Tutsy University of Technology distributed the upper-level control torque to the four hub motors based on the dynamic load transfer of the tires, and designed a sliding mode controller based on slip ratio to correct the torque. Simulation results showed that the lateral acceleration of the vehicle decreased, and the yaw rate could track the reference value well. At the same time, the sideslip angle of the vehicle's center of gravity decreased, and the vehicle's handling stability was further improved. Laghmara H et al. from Sorbonne University distributed the longitudinal torque and yaw torque of the upper-level control according to the ratio of the vertical load transfer of the tires, improving the vehicle's ability to track the desired motion state. Ren B et al. from Jilin University used the torque of the four hub motors as the control variable of the MPC control system. Its control objective function not only considered the performance of the vehicle state tracking the desired state, but also the maximum torque of the hub motors, the tire slip ratio, the torque change, and the vehicle's steering requirements. Computer simulation results show that reasonable torque distribution leads to good controller performance, meeting the driver's expectations. Shi Y and Yu F et al. from Shanghai Jiao Tong University neglected the influence of tire lateral force on yaw moment, assuming that the desired yaw moment is generated by tire longitudinal force. Torque distribution to a single wheel can be viewed as a nonlinear constrained optimization problem. Minimizing energy loss and longitudinal control error is chosen as the objective function, and the problem is solved using Sequence Quadratic Programming (SQP). Simulation verification combined with upper-level control shows improved vehicle handling stability under typical lane-changing conditions and reduced driver load. Optimal control algorithms are currently the preferred choice for researchers studying hub motor torque distribution control. Distributed electric drive vehicles belong to actuator redundancy systems, which can be transformed into an optimization problem under constraints. Using an optimal allocation algorithm can distribute lateral force and yaw moment to the actuators at optimal values, fully utilizing the actuators' capabilities. Hu J et al. from Chongqing University used the highest efficiency of the in-wheel motor as the objective function, and the maximum torque of the in-wheel motor and the drive distribution coefficient of the front wheels as constraints to allocate the target control force and torque. This control algorithm can achieve a maximum motor efficiency of 89.49%, which is 0.42% higher than the efficiency of average distribution. Wu Dongmei from Jilin University used energy saving and safety as the goals of integrated vehicle dynamics control. She used an optimal allocation algorithm to comprehensively distribute torque, which not only improved the stability of vehicle steering and driving, but also improved the efficiency of the in-wheel motor.

[0004] Average load distribution and dynamic load distribution are direct distribution control methods, which have low computational complexity and fast response, but they neglect the constraints of the actuators, failing to fully utilize their capabilities and degrading the overall system performance. SQP, based on mathematical programming algorithms, can fully consider the constraints of the control system and has high accuracy, but the algorithm is relatively complex. Furthermore, in hub motor torque distribution control, the constraint optimization problem fails to fully consider the actuator constraints, and most studies focus on the front and rear axles. The actual driving environment of hub motor-driven intelligent vehicles is complex and variable. Optimal torque distribution must not only satisfy the vehicle's handling stability but also consider energy loss and driving range. Therefore, this invention designs a torque distribution control method for four-wheel hub motor-driven intelligent vehicles, achieving an optimal distribution algorithm that simultaneously considers vehicle stability and economy. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a torque distribution control method for a four-wheel hub motor driven intelligent vehicle. The method applies extension control theory to the torque distribution control of a four-wheel hub motor driven intelligent vehicle, and dynamically allocates weight coefficients based on energy saving and stability according to the driving state of the vehicle, so as to ensure energy saving and stable driving of the vehicle.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] A torque distribution control method for a four-wheel hub motor driven intelligent vehicle, comprising the following specific steps:

[0008] 1) First, establish a two-degree-of-freedom nonlinear dynamic model of the whole vehicle based on the vehicle's own parameters and the input parameters of the working conditions. Then, use the two-degree-of-freedom dynamic model to draw a phase plane diagram of the vehicle's center of gravity sideslip angle and sideslip angular velocity. The horizontal axis is the center of gravity sideslip angle and the vertical axis is the center of gravity sideslip angular velocity.

[0009] 2) Based on the actual sideslip angle, sideslip angular velocity and stable point of the vehicle under driving conditions, the five-parameter method is used to determine the rhombic stable region. The boundary of the rhombic stable region is obtained according to the conditions satisfied by the boundary line of the rhombic stable region.

[0010] 3) Based on the relationship that the four lines of the rhombus boundary are tangent to the boundary of the inner ellipse, the objective function is to maximize the area of ​​the inscribed ellipse. The ranges of the center, the major semi-axis, and the minor semi-axis are determined as constraints. The inscribed ellipse is obtained using the fmincon function of Matlab, which is the classical domain of the extended set.

[0011] 4) Based on the fact that the intersection points of the four lines of the rhomboid boundary coincide with the outer ellipse boundary, use Matlab to solve the system of equations consisting of the equations of the four lines and the equation of the circumscribed ellipse, and obtain the circumscribed ellipse, which is the extensible field of the extensible set.

[0012] 5) Select the sideslip angle and sideslip angular velocity of the vehicle's center of gravity as two characteristic quantities. Based on the extended distance solution method, divide the measurement mode and determine the weighting coefficients.

[0013] 6) Determine the objective functions for torque distribution of the hub motors based on vehicle stability and energy saving. Based on the weighting coefficients established in step 5), obtain the final overall objective function. Set the constraints for achieving the objective function based on the total force, torque and driving constraints required by the vehicle. Solve the optimal solution using Matlab software to obtain the torque of each hub motor.

[0014] Furthermore, step 1) specifically includes:

[0015] 1.1 Based on the aforementioned vehicle parameters, a nonlinear tire model is established using the magic formula, namely:

[0016] According to formula F y =Dsin{Carctan[Bα-E(Bα-arctan(Bα))]} to obtain the tire lateral force;

[0017] Among them, F y Let α be the tire lateral force, α be the tire slip angle, and B, C, D, and E be functions of the tire vertical load and camber angle, which can be obtained by fitting the tire model parameters.

[0018] 1.2 Based on the tire lateral force in step 1.1, establish a two-degree-of-freedom nonlinear dynamic model of the vehicle, namely:

[0019] Ignoring longitudinal forces on the tires and assuming the rear wheel angle is zero, according to the formula... Establish a nonlinear dynamics model for the vehicle;

[0020] Among them, F yf and F yr These represent the lateral forces of the front and rear tires, respectively; m is the total vehicle mass; v is the total vehicle speed; and I... z L represents the rotational inertia of the entire vehicle. f and L r These are the distances from the vehicle's center of gravity to the front and rear axles, respectively; β is the sideslip angle of the vehicle's center of gravity; and γ is the vehicle's yaw rate.

[0021] 1.3 Using the vehicle's center of gravity sideslip angle and sideslip angular velocity as the abscissa and ordinate of the phase plane, respectively, and given different initial values ​​of the vehicle's center of gravity sideslip angle and sideslip angular velocity, by setting different wheel angles, vehicle speeds and road adhesion coefficients, the vehicle's center of gravity sideslip angle-slip angular velocity phase plane diagrams under different driving conditions are obtained.

[0022] Furthermore, step 2) specifically includes:

[0023] 2.1 The five-parameter method is used to determine the rhombic stable boundary, where the stable point is the focus where the phase trajectories converge under different initial states. When the critical phase trajectory converges to the focus, the intersections with the abscissa and ordinate are the left endpoints of the rhombus, respectively. right endpoint Top vertex and lower vertex Connecting endpoints and vertices creates a stable rhombus boundary;

[0024] 2.2 Combine the left endpoint (-β1, 0), right endpoint (β2, 0), and top vertex from step 2.1. and lower vertex The stability domain boundary relation is obtained by substituting the sideslip angle and the angular velocity of the deflection angle. Solve the system of equations (m = 1 to 4) to determine the parameter C of the rhomboid boundary line. 1m C 2m ;

[0025] Furthermore, step 3) specifically includes:

[0026] Based on the tangency relationship between the boundary of the rhombic stable region and the inscribed ellipse, the area of ​​the inscribed ellipse is defined as the objective function:

[0027] J 内 =min(-πR1R2)

[0028] The constraints are:

[0029] Where R1 and R2 are the major and minor axes of the inscribed ellipse, respectively, and a1 and b1 are the x-coordinate and y-coordinate of the center of the inscribed ellipse, respectively.

[0030] The fmincon function provided by Matlab software can be used to solve for the major semi-axis, minor semi-axis, and center position of the inscribed ellipse, thus obtaining the inscribed ellipse of the rhombus, which is the classical field of the extended set.

[0031] Furthermore, step 4) specifically includes:

[0032] According to the formula for the circumscribed ellipse By substituting the two endpoints and two vertices of the boundary of the rhombic stable region into the equations, the major semi-axis, minor semi-axis, and center of the circumscribed ellipse of the rhombic stable region can be obtained. The circumscribed ellipse of the rhombus is determined, which is the extensible domain of the extensible set. The area outside the extensible domain is the non-domain, and the extensible set is finally determined.

[0033] Among them, R a and R b O1 and O2 are the major and minor axes of the circumscribed ellipse, respectively, and O1 and O2 are the x-coordinate and y-coordinate of the center of the circumscribed ellipse, respectively.

[0034] Furthermore, step 5) specifically includes:

[0035] The sideslip angle and sideslip velocity of the vehicle's center of gravity are measured in real time using sensors, thereby obtaining feature points in a two-dimensional feature space. These points are then converted into a one-dimensional extension set, and the extension distance is calculated using the correlation function formula. When the region containing the feature point is determined, and K(s)≥1, η e =1, η s =0; when 0≤K(s)<1, η e =K(s), η s = 1 - K(s); when K(s) < -1, η e =0, η s =1.

[0036] Where P is the characteristic point of the centroid sideslip angle and sideslip angular velocity in the extension set, and P1, P2, P3 and P4 are the intersection points of the connection between point P and the center of the circle, and the extension domain boundary and the classical domain boundary, respectively.

[0037] Furthermore, step 6) specifically includes:

[0038] The objective function formulas for torque distribution control of hub motors based on stability and energy saving are as follows:

[0039]

[0040]

[0041] Constraints:

[0042]

[0043] Among them, T xij (i = f ~ r, j = l ~ r) represents the torque of each hub motor, μ is the road adhesion coefficient, and R ij F is the tire rolling radius. zij For the vertical load on the tire, P loss，ij The power loss of each hub motor is expressed as follows: T x The total longitudinal demand torque is given by ΔM, where ΔM is the additional direct yaw moment, and B is the torque. f and B r The front and rear track widths are δ. f For the front wheel steering angle, T max P is the maximum torque of the hub motor. em This represents the power loss when the hub motor's torque is zero.

[0044] Based on the total force, torque, and driving constraints required by the vehicle, the objective function is set to achieve the desired constraints. The optimal solution is obtained by programming in Matlab software to obtain the torque of each wheel hub motor.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] By employing the solution of this invention, the rational distribution of torque among the four-wheel hub motors in a smart car during operation not only reduces energy loss but also ensures stable vehicle operation. This invention selects two characteristic quantities: the sideslip angle and yaw rate of the vehicle's center of gravity. Using a two-degree-of-freedom nonlinear dynamics model of the vehicle, initial values ​​of the sideslip angle and yaw rate are input to construct a phase plane of the sideslip angle and yaw rate, determining the vehicle's stable driving region and extension set. Based on extension control theory, dynamic weighting coefficients are determined for the objective function of hub motor torque distribution control based on stability and energy saving. This allows for real-time dynamic distribution of hub motor torque, balancing the vehicle's overall driving stability and economy. Attached Figure Description

[0047] Figure 1 This is a technical approach for torque distribution control in a four-wheel hub motor driven intelligent vehicle.

[0048] Figure 2 It is a two-degree-of-freedom nonlinear dynamics model of a car.

[0049] Figure 3 It is a phase plane diagram of the car's center of gravity sideslip angle and sideslip angular velocity.

[0050] Figure 4 It is the classical field boundary of an extension set.

[0051] Figure 5 It is the boundary of the extension field of an extension set.

[0052] Figure 6 It is an extension set.

[0053] Figure 7 It is the extension distance transformation in a two-dimensional extension set. Detailed Implementation

[0054] The following is in conjunction with the appendix Figure 1-7 The present invention will be described in detail, such as Figure 1 The diagram shown is a block diagram of a torque distribution control method for a four-wheel hub motor driven intelligent vehicle. The method includes the following steps:

[0055] 1) First, establish a two-degree-of-freedom nonlinear dynamic model of the vehicle based on its own parameters and operating condition input parameters. Then, use this two-degree-of-freedom dynamic model to plot the vehicle's sideslip angle-slip angular velocity phase plane diagram. The horizontal axis represents the sideslip angle, and the vertical axis represents the sideslip angular velocity. Specifically, this includes:

[0056] 1.1 Based on the aforementioned vehicle parameters, a nonlinear tire model is established using the magic formula, namely:

[0057] According to formula F y =Dsin{Carctan[Bα-E(Bα-arctan(Bα))]} to obtain the tire lateral force;

[0058] Among them, F y Let α be the tire lateral force, α be the tire slip angle, and B, C, D, and E be functions of the tire vertical load and camber angle, which can be obtained by fitting the tire model parameters.

[0059] 1.2 Based on the tire lateral force in step 1.1, establish a two-degree-of-freedom nonlinear dynamic model of the vehicle, such as... Figure 2 As shown, that is:

[0060] Neglecting longitudinal forces on the tires and assuming zero rear wheel steering angle, the dynamic equations are obtained based on the formula of the nonlinear dynamics model of the vehicle.

[0061] Among them, F yf and F yr These represent the lateral forces of the front and rear tires, respectively; m is the total vehicle mass; v is the total vehicle speed; and I... z L represents the rotational inertia of the entire vehicle. f and L r These are the distances from the vehicle's center of gravity to the front and rear axles, respectively; β is the sideslip angle of the vehicle's center of gravity; and γ is the vehicle's yaw rate.

[0062] 1.3 Assuming the car maintains a certain longitudinal speed, based on different initial values ​​of the car's center of gravity sideslip angle and yaw rate, the phase plane diagram of the car's center of gravity sideslip angle versus sideslip rate is obtained, as shown below. Figure 3 As shown;

[0063] Furthermore, step 4) specifically includes:

[0064] According to the formula for the circumscribed ellipse By substituting the two endpoints and two vertices of the boundary of the rhombic stable region, solving the system of equations yields the major semi-axis, minor semi-axis, and center of the circumscribed ellipse of the rhombic stable region. Determining the circumscribed ellipse of the rhombus allows us to define the boundary of the extended domain of the extended set, such as... Figure 5 As shown, the area outside the extensionable field is a non-field, thus ultimately determining the extensionable set, as follows. Figure 6 As shown.

[0065] Among them, R a and R b O1 and O2 are the major and minor axes of the circumscribed ellipse, respectively, and O1 and O2 are the x-coordinate and y-coordinate of the center of the circumscribed ellipse, respectively.

[0066] 2) Based on the actual sideslip angle, sideslip angular velocity, and stable point of the vehicle under driving conditions, the five-parameter method is used to determine the rhombic stable region. The boundary of the rhombic stable region is obtained based on the conditions satisfied by its boundary line, specifically including:

[0067] 2.1 The five-parameter method is used to determine the rhombic stable boundary, where the stable point is the focus where the phase trajectories converge under different initial states. When the critical phase trajectory converges to the focus, the intersections with the abscissa and ordinate are the left endpoints of the rhombus, respectively. right endpoint Top vertex and lower vertex Connecting the endpoints and vertices creates a rhombus-shaped stable domain boundary, such as... Figure 3 As shown;

[0068] 2.2 Take the left endpoint from step 2.2 right endpoint Top vertex and lower vertex The stability domain boundary relation is obtained by substituting the sideslip angle and the angular velocity of the deflection angle. Solve the system of equations (m = 1 to 4) to determine the parameter C of the rhomboid boundary line. 1m C 2m ;

[0069] 3) Based on the relationship that the four lines of the rhombus boundary are tangent to the inner ellipse boundary, the objective function is to maximize the area of ​​the inscribed ellipse. The ranges of the center, major semi-axis, and minor semi-axis are determined as constraints. The inscribed ellipse is obtained using the `fmincon` function in Matlab, thus allowing the classical domain boundary of the set to be extended. Figure 5 As shown, it specifically includes:

[0070] Based on the tangency relationship between the boundary of the rhombic stable region and the inscribed ellipse, the area of ​​the inscribed ellipse is defined as the objective function:

[0071] J 内 =min(-πR1R2)

[0072] The constraints are:

[0073] Where R1 and R2 are the major and minor axes of the inscribed ellipse, respectively, and a1 and b1 are the x-coordinate and y-coordinate of the center of the inscribed ellipse, respectively.

[0074] The fmincon function provided by Matlab software can be used to solve for the semi-major axis, semi-minor axis, and center position of the inscribed ellipse, thus obtaining the inscribed ellipse of the rhombus. This allows for the expansion of the classical domain boundary of the set, such as... Figure 4 As shown;

[0075] 4) Based on the fact that the intersection points of the four lines of the rhomboid boundary coincide with the outer ellipse boundary, use Matlab to solve the system of equations about the equations of the four lines and the equation of the circumscribed ellipse to obtain the circumscribed ellipse, which is the boundary of the extensional domain of the extensional set. The area outside the extensional domain boundary is a non-domain. The extensional set includes classical domains, extensional domains and non-domains.

[0076] 5) Selecting the vehicle's sideslip angle and sideslip angular velocity as two characteristic quantities, and based on the extended distance solution method, dividing the measurement mode and determining the weighting coefficients, specifically including:

[0077] Using sensors to measure the vehicle's center of gravity sideslip angle and sideslip angular velocity in real time, feature points in a two-dimensional feature space are obtained and converted into a one-dimensional extension set, such as... Figure 7 As shown, the extension distance is obtained using the correlation function formula. When the region containing the feature point is determined, and K(s)≥1, η e =1, η s =0; when 0≤K(s)<1, η e =K(s), η s = 1 - K(s); when K(s) < -1, η e =0, η s =1.

[0078] Where P is the characteristic point of the centroid sideslip angle and sideslip angular velocity in the extension set, and P1, P2, P3 and P4 are the intersection points of the connection between point P and the center of the circle and the extension domain boundary and the classical domain boundary, respectively.

[0079] 6) Determine the objective functions for hub motor torque distribution based on vehicle stability and energy saving. Based on the weighting coefficients established in step 5), obtain the overall objective function. Set constraints for achieving the objective function based on the total force, torque, and driving constraints of the vehicle. Solve for optimality using Matlab software to obtain the torque of each hub motor, specifically including:

[0080] The objective function formulas for torque distribution control of hub motors based on stability and energy saving are as follows:

[0081]

[0082]

[0083] Constraints:

[0084]

[0085] Among them, T xij (i = f ~ r, j = l ~ r) represents the torque of each hub motor, μ is the road adhesion coefficient, and R ijF is the tire rolling radius. zij For the vertical load on the tire, P loss,ij The power loss of each hub motor is expressed as follows: Tx is the total longitudinal demand torque, ΔM is the additional direct yaw moment, and B f and B r The front and rear track widths are δ. f For the front wheel steering angle, T max P is the maximum torque of the hub motor. em This represents the power loss when the hub motor's torque is zero.

[0086] Based on the total force, torque, and driving constraints required by the vehicle, the objective function is set to achieve the desired constraints. The optimal solution is obtained by programming in Matlab software to obtain the torque of each wheel hub motor.

Claims

1. A torque distribution control method for a four-wheel hub motor driven intelligent vehicle, comprising the following steps: 1) First, establish a two-degree-of-freedom nonlinear dynamic model of the whole vehicle based on the vehicle's own parameters and the input parameters of the working conditions. Then, use the two-degree-of-freedom dynamic model to draw a phase plane diagram of the vehicle's center of gravity sideslip angle and sideslip angular velocity. The horizontal axis is the center of gravity sideslip angle and the vertical axis is the center of gravity sideslip angular velocity. 2) Based on the actual sideslip angle, sideslip angular velocity and stable point of the vehicle under driving conditions, the five-parameter method is used to determine the rhombic stable region. The boundary of the rhombic stable region is obtained according to the conditions satisfied by the boundary line of the rhombic stable region. 3) Based on the relationship that the four lines of the rhombus boundary are tangent to the boundary of the inner ellipse, the objective function is to maximize the area of ​​the inscribed ellipse. The ranges of the center, the major semi-axis, and the minor semi-axis are determined as constraints. The inscribed ellipse is obtained using the fmincon function of Matlab, which is the classical domain of the extended set. 4) Based on the fact that the intersection points of the four lines of the rhomboid boundary coincide with the outer ellipse boundary, use Matlab to solve the system of equations consisting of the equations of the four lines and the equation of the circumscribed ellipse, and obtain the circumscribed ellipse, which is the extensible field of the extensible set. 5) Select the sideslip angle and sideslip angular velocity of the vehicle's center of gravity as two characteristic quantities. Based on the extended distance solution method, divide the measurement mode and determine the weighting coefficients. 6) Determine the objective functions for torque distribution of the hub motors based on vehicle stability and energy saving respectively. Based on the weight coefficients established in step 5), the final objective function is obtained. Set the constraints for achieving the objective function according to the total force, torque and driving constraints of the vehicle. Solve the optimal solution using Matlab software to obtain the torque of each hub motor. Step 4) specifically includes: According to the circumscribed ellipse formula , and the two endpoints and two vertices of the rhombus stable region boundary are brought in, the long semi-axis, the short semi-axis and the center of the circumscribed ellipse of the rhombus stable region are obtained by solving the equation set, the circumscribed ellipse of the rhombus is determined, that is, the extension domain of the extension set, and the non-domain is outside the extension domain. Finally, the extension set is determined. wherein R a and R b are the major and minor semi-axes of the circumscribed ellipse, respectively, and are the horizontal and vertical coordinates of the center of the circumscribed ellipse, respectively, β is the vehicle mass center side slip angle; Step 6) specifically includes: The objective function formulas for torque distribution control of hub motors based on stability and energy saving are as follows: Constraints: in, T xij ( i = f ~ r , j = l ~ r () represents the torque of each hub motor. μ The road surface adhesion coefficient, R ij The tire's rolling radius, F zij For the vertical load on the tire, The power loss of each hub motor is expressed as follows: , For the total vertical demand torque, M For the additional direct yaw moment, B f and B r The front and rear track widths are... δ f For the front wheel steering angle, T max This represents the maximum torque of the hub motor. P em This represents the power loss when the hub motor's torque is zero.

2. The torque distribution control method for a four-wheel hub motor driven intelligent vehicle according to claim 1, characterized in that, Step 1) specifically includes: 1.1 Based on the vehicle parameters, a nonlinear tire model is established using the magic formula, namely: According to the formula Obtain the lateral force of the tire; in, F y This refers to the lateral force of the tire. α This refers to the tire slip angle. B , C , D and E It is a function of the tire's vertical load and camber angle, and can be obtained by fitting the tire model parameters; 1.2 Based on the tire lateral force in step 1.1, establish a two-degree-of-freedom nonlinear dynamic model of the vehicle, namely: Ignoring longitudinal forces on the tires and assuming the rear wheel angle is zero, according to the formula... Establish a nonlinear dynamics model for the vehicle; in, F yf and F yr These are the lateral forces of the front and rear tires, respectively. m For the overall vehicle quality, v For the total vehicle speed, I z The rotational inertia of the entire vehicle. L f and L r These are the distances from the vehicle's center of gravity to the front and rear axles, respectively. γ This refers to the yaw rate of the entire vehicle. 1.3 Using the vehicle's center of gravity sideslip angle and sideslip angular velocity as the abscissa and ordinate of the phase plane, respectively, and given different initial values ​​of the vehicle's center of gravity sideslip angle and sideslip angular velocity, by setting different wheel angles, vehicle speeds and road adhesion coefficients, the vehicle's center of gravity sideslip angle-slip angular velocity phase plane diagrams under different driving conditions are obtained.

3. The torque distribution control method for a four-wheel hub motor driven intelligent vehicle according to claim 2, characterized in that, Step 2) specifically includes: 2.1 The five-parameter method is used to determine the rhombic stable boundary, where the stable point is the focus where the phase trajectories converge under different initial states. When the critical phase trajectory converges to the focus, the intersections with the abscissa and ordinate are the left endpoints of the rhombus, respectively. Right endpoint Upper vertex and lower vertex Connecting the endpoints and vertices creates a stable rhombus boundary; 2.2 The left endpoint (-) in step 2.1 β 1, 0), right endpoint ( β 2, 0), upper vertex and lower vertex The stability domain boundary relation is obtained by substituting the sideslip angle and the angular velocity of the deflection angle. ( m Solve the system of equations (1-4) to determine the parameters of the rhomboid boundary line. C 1m , C 2m .

4. The torque distribution control method for a four-wheel hub motor driven intelligent vehicle according to claim 3, characterized in that, Step 3) specifically includes: Based on the tangency relationship between the boundary of the rhombic stable region and the inscribed ellipse, the area of ​​the inscribed ellipse is defined as the objective function: The constraints are: ; in, R 1 and R 2 are the major and minor semi-axis of the inscribed ellipse, respectively. and These are the x-coordinate and y-coordinate of the center of the inscribed ellipse, respectively. The fmincon function provided by Matlab software can be used to solve for the major semi-axis, minor semi-axis, and center position of the inscribed ellipse, thus obtaining the inscribed ellipse of the rhombus, which is the classical field of the extended set.

5. The torque distribution control method for a four-wheel hub motor driven intelligent vehicle according to claim 1, characterized in that, Step 5) specifically includes: The sideslip angle and sideslip velocity of the vehicle's center of gravity are measured in real time using sensors, thereby obtaining feature points in a two-dimensional feature space. These points are then converted into a one-dimensional extension set, and the extension distance is calculated using the correlation function formula. Determine the region where the feature points are located. hour, , ; hour, , ; hour, , ; in, P These are the characteristic points of the centroid sideslip angle and sideslip angular velocity in the extension set. P 1. P 2. P 3 and P 4 are respectively P The point is the intersection of the connection between the point and the center of the circle, and the extension of both sides to the boundary of the extended field and the boundary of the classical field, respectively.

Images