CN110293853A - Torque distribution method under four motorized wheels electric car steering situation - Google Patents

Abstract

The torque distribution method under the steering condition of four-wheel independent drive electric vehicles belongs to the field of new energy vehicle control. In order to solve the problem of power and stability of FWID-EV under steering, the fuzzy controller uses the fuzzy control algorithm to calculate the vehicle steering angle. , the fuzzy classification of the variation of the rotation angle to obtain the adjusted driving torque; the adjusted driving torque is used to adjust the initial driving torque to obtain the corrected driving torque; the corrected driving torque is used as the input of the optimal torque distribution controller, and the optimal torque distribution control The controller executes an optimal torque distribution algorithm to determine the torque distribution to the four wheels; the effect is to achieve a more realistic wheel torque distribution for cornering.

Description

Torque distribution method of four-wheel independent drive electric vehicle under steering conditions

technical field

The invention belongs to the field of new energy vehicle control, in particular to a torque distribution system and a working method for steering working conditions of a four-wheel independent drive electric vehicle (FWID-EV).

Background technique

The intensification of global energy problems has promoted the development of new energy vehicles. The four-wheel independent drive electric vehicle has become the focus of the new energy vehicle industry because of its characteristics of less pollution, low energy consumption and independent control of wheel torque in terms of structure. However, there will be many safety problems in practical applications, especially in steering conditions, it is easy to occur a series of dangerous situations such as tail flicking and rollover. How to use its unique advantages to distribute torque to improve vehicle safety and reduce traffic accidents Effects are a key area of electric vehicle research.

Reasonable torque distribution is the core technology of FWID-EV control, and it is also a necessary step for the realization of intelligentization and mass production of electric vehicles. The main function of torque distribution is to reasonably distribute the driving torque obtained by the vehicle speed model and the adjustment torque required for driving to the four motors, so that they can meet the driving needs of the vehicle and achieve the best performance of the driving vehicle. At present, in terms of torque distribution, most of the literature and research are using average torque distribution or vertical distribution. Although these two methods of torque distribution are simple in principle, they do not take into account the possible instability during the steering process, so they are not suitable for steering. drive. In the torque distribution of steering driving, the most commonly used distribution method is to distribute the torque according to some reasonable rules. The difficulty of distribution, but did not fully explore the unique advantages of independent control of the four wheels of four-wheel independent drive electric vehicles, nor did it consider the transfer of the center of gravity during the steering process. Therefore, in order to improve the steering stability of FWID-EV and better apply the upper-level control algorithm to the driving of the wheels, it is necessary to develop a torque distribution specifically for steering conditions.

Contents of the invention

In order to solve the problem of power and stability of FWID-EV under steering, the present invention provides a torque distribution method under the steering condition of four-wheel independent drive electric vehicle. The steering angle and the fuzzy classification of the variation of the steering angle are used to obtain the adjusted driving torque; the adjusted driving torque is used to adjust the initial driving torque to obtain the corrected driving torque;

The corrected driving torque is used as the input of the optimal torque distribution controller, and the optimal torque distribution controller executes the optimal torque distribution algorithm to determine the torque distributed to the four wheels;

Wherein, the first item of the objective function of the optimal torque distribution algorithm is the difference between the vehicle demanded yaw moment and the actual yaw moment of the vehicle steering, and the vehicle demanded yaw moment is the corrected driving torque, the target The second term of the function is tire utilization.

Beneficial effects: the present invention uses the fuzzy control algorithm to carry out the fuzzy classification of the vehicle steering angle and the variation of the steering angle under the consideration of the stability of the vehicle and the advantages of the in-wheel motor electric vehicle, and adjusts the driving torque obtained by the closed-loop speed control model of the driver , the output of the fuzzy controller is the corrected drive torque value, which is used as the vehicle demand torque, and the difference between the actual yaw moment of the vehicle during steering and driving is used as the first item of the objective function in the optimal torque distribution controller. Utilization rate is used as the second item of the objective function, and the quadratic programming problem is constructed, and the driving force of the four wheels is distributed by using the method of effective set solution, so as to realize the torque distribution of the vehicle.

Description of drawings

Figure 1 is a diagram of the driver's closed-loop speed model.

Figure 2 is the overall framework of the program.

Detailed ways

In order to ensure the power and stability of the FWID-EV under steering, the present invention provides a torque distribution method under the steering condition of a four-wheel independent drive electric vehicle. In one embodiment, the stability of the vehicle and the wheel hub Advantages of motor electric vehicles, use fuzzy control algorithm to fuzzy classify vehicle steering angle and steering angle variation, obtain adjusted driving torque, and adjust the driving torque obtained by the driver's closed-loop speed control model, and output the corrected driving torque value, which is used as the vehicle demand torque, and the difference between the vehicle’s actual yaw moment during turning and driving is used as the first item of the objective function in the optimal torque distribution controller, and the tire utilization rate is used as the second item of the objective function. For the quadratic programming problem, the driving force of the four wheels is allocated by using the method of effective set solution.

In a nutshell, the moment distribution algorithm is as follows:

1. The overall moment distribution strategy is shown in Figure 2. The vehicle angle signal is obtained according to the gyroscope and optical encoder installed on the rotation angle shaft, and the vehicle speed is obtained according to the speed sensor. The steering wheel angle and speed are used as the input signal of the driver's closed-loop controller. After processing by the PID controller, the driver's closed-loop control The output of the device is the initial drive torque.

2. The input of the fuzzy controller is the steering wheel angle and the change of the steering wheel angle. The controller obtains the yaw moment change required to be corrected under the steering condition through the fuzzy control algorithm, and finally combines with the initial drive torque to obtain the controller output as the correction the driving torque.

3. Input the final corrected drive torque output by the controller in the previous step and the adjustment torque output by the middle controller installed on the FWID-EV car body into the optimal torque distribution controller, and use the quadratic programming algorithm to determine the distribution to the four The torque of the wheels, the output of the controller directly acts on the four wheels, and the trend of the complete steering movement of the vehicle

The following is a detailed description of the above torque distribution algorithm:

A. Driving force correction:

A1. Initial drive torque

When the vehicle is running, the initial driving torque should be obtained according to the driver's demand to maintain the basic motion of the vehicle. In this embodiment, the PID closed-loop feedback model is selected as the ideal vehicle speed for the driver's closed-loop controller to follow the driver's demand in terms of the driver's speed model. The control input is the difference between the actual vehicle speed and the ideal vehicle speed, and the control output is the accelerator opening. Determine the size of the throttle opening through the speed difference and the driver's reaction time, and then connect the throttle operating characteristic table to output the current required output drive torque, and finally distribute the current torque to the four wheels reasonably through the lower distributor. Speed regulation for four-wheel independent drive electric vehicles.

This embodiment mainly distributes and controls the motor torque, so the deceleration and braking effect of the motor is not considered in the driver speed model.

A2. Driving Force Correction

When the vehicle enters the steering state, the driver controls the deceleration and turning by depressing the accelerator pedal and turning the steering wheel. The opening of the throttle in the A1 step driver's speed model determines the initial driving force of the vehicle, and the torque distributor sends commands to the motors installed on the steering wheels according to the designed torque distribution algorithm. During the steering process, the only four wheels that are in contact with the ground are the four wheels. At this time, the two steering wheels will generate two restraint forces on the ground respectively. These two restraint forces generate moments around the left and right axes respectively. The values of these two moments The torque required for steering is formed between the left and right wheels. This steering torque and the operating torque input by the driver jointly resist the frictional torque generated between the parts in the steering system and the inertial moment of the wheels, and drive the vehicle to turn to complete the steering process. The steering operation status of the vehicle will be fed back to the driver through the action of the vehicle dynamics model, allowing the operator to complete the steering operation at the next moment according to the actual state until the vehicle is completely out of the steering condition.

From the above analysis, it can be seen that the running process of the vehicle under the steering condition is a motion process in which the driver's closed-loop driving force, yaw moment and lateral centripetal moment overcome the resistance to steer the vehicle. In the whole process of torque distribution, by adjusting the longitudinal force of each wheel, its effect on the yaw moment will be different. For example, when the vehicle is turning right, if the ideal yaw moment is in a small range, the driving torque of the left front wheel motor can be reduced to avoid oversteering; if the ideal yaw moment is in a large range, it is necessary to The moments of the remaining two wheels are controlled comprehensively. Therefore, in this embodiment, the stability and dynamic performance of the vehicle during steering are improved by correcting the driving torque of the two front wheels.

According to the dynamic model, the torque difference between the left front wheel and the right front wheel can be written as:

In the formula, T _c1 is the driving torque of the left wheel, T _c2 is the driving torque of the right wheel, T _d1 is the steering driving torque of the left wheel, T _d2 is the steering driving torque of the right wheel, F _x1 is the driving force of the left wheel, and F _x2 is The driving force of the right wheel, l _c is the track of the left and right wheels. r is the wheel radius. It can be seen from the above formula that, when the vehicle is turning, if the torque of the outer wheel of the vehicle is greater than the torque of the inner wheel, the original driving torque will be corrected during the entire steering process.

In the formula, T _d is the original driving torque, T ₁ ' is the corrected driving torque of the left front wheel, T ₂ ' is the corrected driving torque of the right front wheel, ΔT _d is the adjusted driving torque, and the driving torque is adjusted by adjusting the size of ΔT _d Adjustment to increase dynamics and stability during steering.

In this embodiment, the fuzzy control algorithm is used to obtain the value of the adjusted driving torque ΔT _d , and the driver's input angle and angle change rate are used as the input of the fuzzy controller during the actual driving of the vehicle. The torque of the left front wheel and the right front wheel of the vehicle The difference is the output, and the fuzzy control rules are established according to the driver's steering experience, and the output is the adjusted driving torque.

Adjusting the driving torque According to formula (2), the original driving torque is adjusted to obtain the corrected driving torque. The adjustment drive torque is obtained for the following scheme.

A2. Adjust the acquisition of driving torque

The acquisition of the driving torque in this embodiment is based on the driving experience to collect the driving angle signal of the four-wheel independent drive electric vehicle, and obtain the value range of the front wheel angle and the rate of change of the angle, and take the domain of the front wheel angle as [- 60, 60], it is stipulated that the wheel is positive when it turns left, and it is negative when it turns right. Set [NB, NM, NS, ZO, PS, PM, PB], where NB means that the front wheel angle is around -60 degrees, NM means that the front wheel angle is between -50 and -30 degrees, and NS means that the front wheel angle is between -30～-10 degrees, ZO means the front wheel angle is between -10～10 degrees, PS means the front wheel angle is between 10～30 degrees, PM means the front wheel angle is between 30～50 degrees, PB means the front wheel The rotation angle is about 60 degrees. Each membership function is a trigonometric function. The range of fuzzy domain of change of front wheel angle is [0, 1], and the subitem of fuzzy language is [FA, MI, SL, ZO], where FA means that the absolute value of front wheel angle change rate is between 0.75 and 1, MI Indicates that the absolute value of the change rate of the front wheel angle is between 0.5 and 0.75, SL indicates that the absolute value of the change rate of the front wheel angle is between 0.25 and 0.5, and ZO indicates that the absolute value of the change rate of the front wheel angle is between 0 and 0.25. The speed change requirement of the steering angle during operation also adopts the triangular membership function. The output of the fuzzy controller is the longitudinal torque difference between the left front wheel and the right front wheel, and its domain is set as [-8080], and its fuzzy language is expressed as [NB, NM, NS, ZO, PS, PM, PB] , the fuzzy rules are shown in Table 1, and thus the adjusted driving torque is obtained.

B. Optimal moment distribution algorithm

After the adjusted drive torque is determined, the drive torque is adjusted to obtain the corrected drive torque, which is combined with the vehicle's controlled adjustment force to determine the actual distribution torque of the four wheels using an optimal distribution algorithm. The optimal torque distribution is a torque distribution method under the comprehensive consideration of road conditions, motor output limitation and axle load transfer.

B1. Optimal moment distribution algorithm objective function

The first item of the objective function: in order to ensure that the torque output from the torque distributor to the four motors meets the torque required by the controller as far as possible under the influence of road conditions and other factors, that is, the actual driving steering torque required by the steering wheel of the vehicle and the vehicle’s demand The difference between the applied theoretical yaw moments should be as small as possible. The vehicle dynamics model is analyzed to obtain:

The longitudinal motion equation of the car along the x-axis:

The yaw motion equation of the car around the center of mass:

Write the two equations (3) and (4) in matrix form, and the mathematical expression is as follows:

V = Bu (5)

in

m is the weight of the vehicle, v _x is the longitudinal velocity of the vehicle, v _y is the lateral velocity of the vehicle, γ is the yaw rate of the vehicle, β is the sideslip angle of the vehicle center of mass, F _x1 is the longitudinal force of the left front wheel of the vehicle, F _x2 is the vehicle The longitudinal force of the right front wheel, F _x3 is the longitudinal force of the left rear wheel of the vehicle, F _x4 is the longitudinal force of the right rear wheel of the vehicle, F _y1 is the lateral force of the left front wheel of the vehicle, F _y2 is the lateral force of the right front wheel of the vehicle, F _y3 is the lateral force of the left rear wheel of the vehicle, F _y4 is the lateral force of the right rear wheel of the vehicle, δ _f is the rotation angle of the front wheel, I _z is the moment of inertia of the vehicle around the Z axis, M _x is the yaw moment, l _f is the center of gravity of the vehicle to The distance from the front axle, l _r is the distance from the center of gravity of the vehicle to the rear axle, l _w is the wheel base, Fx is the driving force of the wheel, T ₃ is the driving torque of the left rear wheel, T ₄ is the driving torque of the right rear wheel;

The first part of the objective function is used for the operation control under the steering of the vehicle. The purpose is to make the actual steering situation of the vehicle as similar as possible to the steering situation expected by the driver. Therefore, the distribution error is taken as the first control target of the torque distribution controller, and the upper Written in matrix norm form:

The second item of the objective function: During the steering process, the stability of the vehicle is the most important part of the entire controller design, so the concept of tire utilization is introduced here. The tire utilization refers to the ground reaction force that the tire receives and its occurrence The ratio of the ultimate force before locking is usually used to reflect the stability of the vehicle, which can be expressed by the following formula.

Among them, μ is the road surface adhesion coefficient, η _i is the tire utilization rate, F _xi is the wheel longitudinal force, F _yi is the wheel lateral force, F _zi is the wheel vertical force, i=1, 2, 3, 4 are respectively the left front wheel, right front wheel, left rear wheel and right rear wheel.

The stability of the car can be expressed by the utilization rate of the tires. The purpose of setting it as the second objective function is to make the adhesion utilization rate of the four tires of the vehicle as small as possible. The higher the tire utilization rate, the worse the stability of the vehicle. , the limit value of the tire utilization rate is 1, at this time the stability reaches the extreme value, and the vehicle will fall into an unstable state. Therefore, in order to ensure the safety of the vehicle, η _i should be reduced under controllable conditions. This article mainly considers the improvement of stability through the optimal distribution of torque, so the effect of lateral force is ignored, and the second objective function is the sum of four tire utilization ratios:

Also write it in matrix norm form:

Among them, T _xi ' is the corrected driving torque, W _u is the weight matrix of the driving torque matrix u, which is used to determine the proportion relationship between the elements in u,

When using quadratic programming to solve the objective function, in order to minimize ||Bu-V||, the weight coefficient ξ is introduced here, and the final objective function can be written as:

Among them, F _z1 is the vertical force of the left front wheel of the vehicle, F _z2 is the vertical force of the right front wheel of the vehicle, F _z3 is the vertical force of the left rear wheel of the vehicle, and F _z4 is the vertical force of the right rear wheel of the vehicle. b is the linear matrix of quadratic programming, A is the weight matrix of quadratic programming, and V is the matrix of longitudinal force and yaw moment. W _v is the weight matrix of the distribution error, which is used to determine the weight relationship among the elements in the first distribution error, and the weight coefficient ξ is taken as 10 ⁶ .

B2. Constraints of Optimal Moment Distribution Algorithm

The first item of the constraint condition of the objective function: when the wheel slips, the lateral force and longitudinal adhesion between the wheel and the ground will be greatly reduced, and the vehicle will be in a very dangerous situation. In order to ensure the safety of the vehicle, prevent the wheel from slipping The risk of excessive slip occurs during driving. This paper first considers the slip ratio adjustment torque of the four wheels as the first constraint condition, that is, as long as the actual absolute value of the slip ratio of a certain wheel is greater than the set maximum absolute value of the slip ratio , then the torque of this wheel will be directly determined by the slip rate adjustment torque, which can be transformed into the mathematical form of the equation constraint in the quadratic programming problem:

Among them, s _i is 1 or 0 (i=1,2,3,4), if the slip rate is greater than 0.2, s _i is set to 1, if not greater than s _i is set to 0, T _s1 is the slip rate adjustment of the left front wheel Moment, T _s2 is the right front wheel slip rate adjustment torque, T _s3 is the left rear wheel slip rate torque, T _s4 is the right rear wheel slip rate torque.

The second item of constraints: through the construction and analysis of the motor model, it can be seen that the permanent magnet brushless DC motor can meet the working characteristics and motion requirements of the vehicle. The torque of the DC motor under different speed conditions is different. The torque and speed characteristics of the motor should be considered when constraining. The relationship between the power, speed and torque of the motor can be expressed as follows.

Among them, T _N is the maximum value of torque, n _N is the dividing point between the constant torque and constant power speed of the motor, n is the wheel speed, n _N is the maximum value of the motor speed, and P _N is the maximum power of the motor.

This embodiment mainly studies the torque distribution of the four wheels, so the influence of factors such as battery loss and charging coefficient on the motor is ignored, so the inequality constraints in the quadratic programming problem are expressed as:

T _brmax ≤ T _i ≤ T _drmax

Among them, T _brmax is the maximum value of the braking torque that the motor can give, T _i is the torque finally distributed to the four motors by the torque distributor, and T _drmax is the minimum value of the driving torque that the motor can give.

Constraint objective third term: The moment divider takes tire adhesion as the second objective function, so the saturation limit of tire force is considered when considering constraints. During the operation of the vehicle, under the action of the vertical load of the vehicle, lateral force and tangential force will be generated between the tire and the ground. When the side slip angle of a certain tire is fixed, the lateral force will gradually increase with the increasing driving force When the driving force increases to a certain limit, the lateral force tends to 0, and the adhesion is almost completely occupied by the tangential force. At this time, the tire cannot provide more lateral force, and the lateral adhesion of the tire will become very limited. This rule also occurs when the tire acting force is the braking force. The circle with the maximum adhesion between the tire and the ground is called the friction circle. The lateral force and tangential force of the tire are distributed on the friction circle perpendicular to each other, and can only change within the friction circle. In actual driving, the friction circle is usually an ellipse. In addition, the longitudinal force is also limited by the tire condition. Combining the constraints of the tire condition and the road surface attachment ellipse, the feasible region of the tire force can be obtained. During the steering process, the car is not only constrained by the driving force, braking force and steering force, but also limited by the friction force of the road surface. The vector sum of the four constraints synthesizes the longitudinal tire force and the lateral tire force into an approximate The ellipse, so when the vehicle is running, the tire force must be within the attachment ellipse, which can be written as a mathematical expression to obtain the constraint conditions of the tire force as the following formula.

sort it out

When the vehicle is running, the tire longitudinal force is limited by the maximum torque of the motor:

Among them, T _xi is the wheel longitudinal moment.

Arrange the above three constraints, and finally form the constraints of the quadratic programming problem

B3. Optimal moment distribution algorithm solution method

According to B1 and B2, the standard formula of quadratic programming can be obtained as follows:

Equation 17 can be solved by the effective set method. The biggest difficulty of the effective set method is that the effective set is not known, so a geometric sequence is constructed to approximate it. The solution process is as follows:

Among them, u _min is the matrix of the minimum value of the driving torque, and u _max is the matrix of the maximum value of the driving torque.

Write the inequalities in the constraints in matrix form

Select the starting point u ₀ satisfying Equation 18, use W to denote the effective constraint index set at this point, set u _k (k=0,1,2,...,N-1) to search along the d _k direction , where N is the number of searches for the kth time, and d _k is the step size of the descent along the gradient minimum direction. Taking the two constraint indicators in Equation 18 as equality constraints, the above equation can be written as:

Among them, u _k is the u value of the kth search, and d _i is the search step size.

If u _k + d _i is a feasible solution to the problem, then the step size α _k = 1, u _{k + 1} = u _k + d _i solve the Lagrangian multiplier according to the following formula

If the multiplier λ≥0, then the optimal solution is u _k+1 . If λ<0, the constraint condition corresponding to the minimum λ is removed from the constraint index set W, and the next iteration of u _k is performed.

If u _k + d _i is not a feasible solution to the problem, use the following formula to find the maximum step size α _k that satisfies the feasible condition, and modify the effective set sequence.

α _k ＝max{α _k ∈[0,1]:u _min ≤u _k +α _k d _i ≤u _max } (21)

Let u _k+1 ＝u _k +α _k d _i , add this iteration point and the current effective constraint into the working set to start the next update, and so on, and finally find the best point in the feasible region.

Compared with the prior art, this embodiment has the following beneficial effects:

1 In this embodiment, an optimal torque distribution algorithm for steering conditions is designed. The driving force is corrected and the structure of the quadratic programming problem not only ensures the stability of the vehicle during the steering process, but also ensures the stability of the vehicle during driving. power.

3 This embodiment uses the fuzzy control algorithm, comprehensively considers safety and dynamics according to the driver's experience, and adds a slip rate constraint to the final optimal torque distribution constraint to avoid the risk of vehicle slippage, so that the driving force correction is designed. Strategy and Optimal Moment Distribution System.

4 This embodiment designs a torque distribution strategy that considers the danger of understeer and oversteer in the steering process of FWID-EV. This strategy introduces the method of changing the driving force and considers the tire saturation condition in the objective function , load transfer and slipping hazards. Finally, the effective set method is used to solve the structural problems to improve the steering stability of FWID-EV and ensure its dynamic performance, which can effectively reduce the probability of danger during the steering process of the vehicle.

In this embodiment, the combination of the fuzzy algorithm and the optimal torque distribution algorithm can greatly improve the stability of the steering condition while ensuring the dynamic performance of the FWID-EV, so as to reduce traffic accidents.

In one embodiment: a torque distribution method under the steering condition of a four-wheel independently driven electric vehicle:

The fuzzy controller uses the fuzzy control algorithm to carry out fuzzy classification of the steering angle and the variation of the steering angle of the vehicle to obtain the adjusted driving torque; the adjusted driving torque is used to adjust the initial driving torque to obtain the corrected driving torque;

The corrected driving torque is used as the input of the optimal torque distribution controller, and the optimal torque distribution controller executes the optimal torque distribution algorithm to determine the torque distributed to the four wheels;

Wherein, the first item of the objective function of the optimal torque distribution algorithm is the difference between the vehicle demanded yaw moment and the actual yaw moment of the vehicle steering, and the vehicle demanded yaw moment is the corrected driving torque, the target The second term of the function is tire utilization.

Further, the execution steps of the optimal moment distribution algorithm are:

Construct the objective function,

create constraints,

Solve and perform optimal moment distribution.

Further, the first item of the constraint condition is the slip ratio of the four wheels,

The second term of the constraints is the torque-speed characteristic of the motor,

The third term of the constraint is the tire force.

Furthermore, the standard formula of quadratic programming is obtained from the objective function and constraints, and it is solved by the effective set method, and a geometric sequence is constructed to approximate it during the solution.

Further, the driver's speed model uses the PID closed-loop feedback model as the driver's closed-loop controller to follow the ideal vehicle speed demanded by the driver. The control input is the difference between the actual vehicle speed and the ideal vehicle speed, and the control output is the accelerator opening. Through the speed difference Value and the driver's reaction time determine the throttle opening, connect the throttle operating characteristic table, and output the current required output driving torque as the initial torque.

Further, the fuzzy controller uses the fuzzy control algorithm to carry out fuzzy classification of the vehicle steering angle and the amount of change in the steering angle, and the method to obtain the adjustment of the driving torque is:

Collect the driving angle signal of the four-wheel independently driven electric vehicle to obtain the value range of the front wheel angle and the rate of change of the front wheel angle. The domain of the front wheel angle is taken as [-60, 60], and when the wheels turn left is positive, and is negative when turning right;

The fuzzy domain of the front wheel rotation angle is [-60, -40, -20, 0, 20, 40, 60], which is defined as 7 fuzzy subsets [NB, NM, NS, ZO, PS, PM, PB ], NB means the front wheel angle is around -60 degrees, NM means the front wheel angle is between -50～-30 degrees, NS means the front wheel angle is between -30～-10 degrees, ZO means the front wheel angle is between -10～ Between 10 degrees, PS indicates that the front wheel rotation angle is between 10 and 30 degrees, PM indicates that the front wheel rotation angle is between 30 and 50 degrees, and PB indicates that the front wheel rotation angle is about 60 degrees;

The range of fuzzy discourse domain of front wheel angle change rate is [0, 1], and the fuzzy language sub-item is defined as [FA, MI, SL, ZO], FA means that the absolute value of front wheel angle change rate is between 0.75 and 1, MI Indicates that the absolute value of the change rate of the front wheel angle is between 0.5 and 0.75, SL indicates that the absolute value of the change rate of the front wheel angle is between 0.25 and 0.5, and ZO indicates that the absolute value of the change rate of the front wheel angle is between 0 and 0.25;

The output of the fuzzy controller is the longitudinal torque difference between the left front wheel and the right front wheel, and its domain is set as [-8080], and its fuzzy language is expressed as [NB, NM, NS, ZO, PS, PM, PB] , the fuzzy rules are shown in the table below;

The driving torque is adjusted by fuzzy rules.

Furthermore, the torque difference between the left front wheel and the right front wheel of the car is:

In the formula: T _c1 is the driving torque of the left wheel, T _c2 is the driving torque of the right wheel, T _d1 is the steering driving torque of the left wheel, T _d2 is the steering driving torque of the right wheel, F _x1 is the driving force of the left wheel, F _x2 is The driving force of the right wheel, lc is the track of the left and right wheels, _r is the radius of the wheel;

When the car is turning, if the torque of the outer wheel of the car is greater than the torque of the inner wheel, the original driving torque is corrected according to the following formula during the entire steering process to obtain the corrected driving torque:

In the formula:

T _d is the original driving torque, T ₁ ′ is the corrected driving torque of the left front wheel, T ₂ ′ is the corrected driving torque of the right front wheel, and ΔT _d is the adjusted driving torque.

Further, the objective function is:

Where: W _u is the weight matrix of the driving moment matrix u; u is the driving moment matrix, u=[T ₁ ', T ₂ ', T ₃ , T ₄ ] ^T ; ξ is the weight coefficient; W _v is the weight matrix for assigning errors;

b is the linear matrix of quadratic programming; A is the weight matrix of quadratic programming; F _x is the driving force of the wheel, M _x is the yaw moment; δ _f is the front wheel angle, r is the radius of the wheel, l _w is the wheel wheel Distance, l _f is the distance from the center of gravity of the vehicle to the front axle.

Further, the constraints described are:

The first item of the constraint condition of the objective function: the slip rate adjustment torque of the four wheels;

Transform this into the mathematical form of the equality constraint in a quadratic programming problem:

Among them, s _i is 1 or 0 (i=1,2,3,4), if the slip rate is greater than 0.2, s _i is set to 1, if not greater than s _i is set to 0, T _s1 is the slip rate adjustment of the left front wheel Moment, T _s2 is the slip rate adjustment torque of the right front wheel, T _s3 is the slip rate torque of the left rear wheel, and T _s4 is the slip rate torque of the right rear wheel;

The second item of the constraint condition of the objective function: the torque and speed characteristics of the motor;

Transform this into the mathematical form of the equality constraint in a quadratic programming problem:

T _brmax ≤ T _i ≤ T _drmax

Among them, T _brmax is the maximum value of the braking torque that the motor can give, T _i is the torque finally distributed to the four motors by the torque distributor, and T _drmax is the minimum value of the driving torque that the motor can give.

The third item of the constraint condition of the objective function: tire force;

Transform this into the mathematical form of the equality constraint in a quadratic programming problem:

Among them, i=1,2,3,4, which are left front wheel, right front wheel, left rear wheel and right rear wheel respectively, r is wheel radius, μ is road surface adhesion coefficient, T _xi is wheel longitudinal moment, F _yi is the wheel lateral force, F _zi is the wheel vertical force;

The three constraints are sorted out to form the constraints of the quadratic programming problem:

Further, the solution and optimal moment distribution are carried out:

The standard formula of quadratic programming is as follows:

Where: W _u is the weight matrix of the driving moment matrix u; u is the driving moment matrix, u=[T ₁ ', T ₂ ', T ₃ , T ₄ ] ^T ; ξ is the weight coefficient; W _v is the weight matrix for assigning errors;

b is the linear matrix of quadratic programming; A is the weight matrix of quadratic programming; F _x is the driving force of the wheel, M _x is the yaw moment; δ _f is the front wheel angle, r is the radius of the wheel, l _w is the wheel wheel distance, l _f is the distance from the center of gravity of the vehicle to the front axle; u _min is the matrix of the minimum value of the driving torque, and u _max is the matrix of the maximum value of the driving torque.

Construct a geometric sequence to approximate equation (17), so that it can be solved by the effective set method, and the solution process is as follows:

Write the inequalities in the constraints in matrix form

Select the starting point u ₀ satisfying formula (18), use W to denote the effective constraint index set at this point, set u _k (k=0,1,2,...,N-1) along d _k Direction search, where N is the number of searches for the kth time, and d _k is the step size that descends along the direction of the minimum gradient. The two constraint indicators in formula (18) are regarded as equality constraints, and formula (18) is written as:

Among them, u _k is the u value of the kth search, and d _i is the search step size;

If u _k + d _i is a feasible solution to the problem, then the step size α _k = 1, u _k+1 = u _k + d _i , and the Lagrangian multiplier can be obtained according to the following formula

If the multiplier λ≥0, the optimal solution is u _k+1 , if λ<0, the constraint condition corresponding to the minimum λ is removed from the constraint index set W, and the next iteration of u _k is performed;

If u _k + d _i is not a feasible solution to the problem, use formula (21) to find the maximum step size α _k that satisfies the feasible condition, and correct the effective set sequence;

α _k ＝max{α _k ∈[0,1]:u _min ≤u _k +α _k d _i ≤u _max } (21)

Let u _k+1 ＝u _k +α _k d _i , add this iteration point and the current effective constraint into the working set to start the next update, and so on, and finally find the best point in the feasible region.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto, any person familiar with the technical field within the technical scope of the disclosure of the present invention, according to the present invention Any equivalent replacement or change of the created technical solution and its inventive concept shall be covered within the scope of protection of the present invention.

Claims (10)

1. a method for moment distribution under the steering condition of a four-wheel independently driven electric vehicle, characterized in that: The fuzzy controller uses the fuzzy control algorithm to carry out fuzzy classification of the vehicle steering angle and the variation of the steering angle to obtain the adjusted driving torque; the adjusted driving torque is used to adjust the initial driving torque to obtain the corrected driving torque; The corrected driving torque is used as the input of the optimal torque distribution controller, and the optimal torque distribution controller executes the optimal torque distribution algorithm to determine the torque distributed to the four wheels; Wherein, the first item of the objective function of the optimal torque distribution algorithm is the difference between the vehicle demanded yaw moment and the actual yaw moment of the vehicle steering, and the vehicle demanded yaw moment is the corrected driving torque, the target The second term of the function is tire utilization. 2. the torque distribution method under the four-wheel independently driven electric vehicle steering working condition as claimed in claim 1, is characterized in that: the execution step of optimal torque distribution algorithm is: Construct the objective function, create constraints, Solve and perform optimal moment distribution. 3. The moment distribution method under the four-wheel independently driven electric vehicle steering working condition as claimed in claim 2, is characterized in that: The first term of the constraints is the slip ratio of the four wheels, The second term of the constraints is the torque-speed characteristic of the motor, The third term of the constraint is the tire force. 4. as claimed in claim 2, the moment distribution method under the four-wheel independently driven electric vehicle steering mode, is characterized in that: obtain the quadratic programming standard formula by objective function and constraint condition, solve it with effective set method, And construct a geometric sequence to approximate it in the solution. 5. The moment distribution method under the steering working condition of four-wheel independent drive electric vehicle as claimed in claim 1, is characterized in that: the driver's speed model selects the PID closed-loop feedback model as the driver's closed-loop controller, following the ideal of driver's demand. Vehicle speed, the control input is the difference between the actual vehicle speed and the ideal vehicle speed, the control output is the throttle opening, the throttle opening is determined by the speed difference and the driver's reaction time, connected to the throttle operating characteristic table, and the current required output drive torque is output is the initial torque. 6. The torque distribution method under the steering working conditions of four-wheel independent drive electric vehicles as claimed in claim 1, characterized in that: the fuzzy controller performs fuzzy classification of the vehicle steering angle and the amount of change in the steering angle with a fuzzy control algorithm, and obtains the adjustment drive The method of moment is: Collect the driving angle signal of the four-wheel independently driven electric vehicle to obtain the value range of the front wheel angle and the rate of change of the front wheel angle. The domain of the front wheel angle is taken as [-60, 60], and when the wheels turn left is positive, and is negative when turning right; The fuzzy domain of the front wheel rotation angle is [-60, -40, -20, 0, 20, 40, 60], which is defined as 7 fuzzy subsets [NB, NM, NS, ZO, PS, PM, PB ], NB means that the front wheel angle is around -60 degrees, NM means that the front wheel angle is between -50 and -30 degrees, NS means that the front wheel angle is between -30 and -10 degrees, ZO means that the front wheel angle is between -10 and Between 10 degrees, PS indicates that the front wheel rotation angle is between 10 and 30 degrees, PM indicates that the front wheel rotation angle is between 30 and 50 degrees, and PB indicates that the front wheel rotation angle is about 60 degrees; The range of fuzzy domain of change rate of front wheel angle is [0, 1], and the subitem of fuzzy language is defined as [FA, MI, SL, ZO]. FA means that the absolute value of front wheel angle change rate is between 0.75 and 1, and MI Indicates that the absolute value of the change rate of the front wheel angle is between 0.5 and 0.75, SL indicates that the absolute value of the change rate of the front wheel angle is between 0.25 and 0.5, and ZO indicates that the absolute value of the change rate of the front wheel angle is between 0 and 0.25; The output of the fuzzy controller is the longitudinal torque difference between the left front wheel and the right front wheel, and its domain is set as [-8080], and its fuzzy language is expressed as [NB, NM, NS, ZO, PS, PM, PB] , the fuzzy rules are shown in the table below; The driving torque is adjusted by fuzzy rules. 7. The torque distribution method under the steering working condition of four-wheel independently driven electric vehicles as claimed in claim 6, wherein the torque difference between the left front wheel and the right front wheel of the vehicle is: In the formula: T _c1 is the driving torque of the left wheel, T _c2 is the driving torque of the right wheel, T _d1 is the steering driving torque of the left wheel, T _d2 is the steering driving torque of the right wheel, F _x1 is the driving force of the left wheel, F _x2 is The driving force of the right wheel, lc is the track of the left and right wheels, _r is the radius of the wheel; When the car is turning, if the torque of the outer wheel of the car is greater than the torque of the inner wheel, the original driving torque is corrected according to the following formula during the entire steering process to obtain the corrected driving torque: In the formula: T _d is the original driving torque, T ₁ ′ is the corrected driving torque of the left front wheel, T ₂ ′ is the corrected driving torque of the right front wheel, and ΔT _d is the adjusted driving torque. 8. The moment distribution method under the steering working condition of four-wheel independent drive electric vehicle as claimed in claim 2, is characterized in that: described objective function is: Where: W _u is the weight matrix of the driving moment matrix u; u is the driving moment matrix, u=[T ₁ ', T ₂ ', T ₃ , T ₄ ] ^T ; ξ is the weight coefficient; W _v is the weight matrix for assigning errors; b is the linear matrix of quadratic programming; A is the weight matrix of quadratic programming; F _x is the driving force of the wheel, M _x is the yaw moment; δ _f is the front wheel angle, r is the radius of the wheel, l _w is the wheel wheel Distance, l _f is the distance from the center of gravity of the vehicle to the front axle. 9. The moment distribution method under the steering condition of four-wheel independent drive electric vehicle as claimed in claim 8, characterized in that: the constraint conditions are: The first item of the constraint condition of the objective function: the slip rate adjustment torque of the four wheels; Transform this into the mathematical form of the equality constraint in a quadratic programming problem: Among them, s _i is 1 or 0 (i=1,2,3,4), if the slip rate is greater than 0.2, s _i is set to 1, if not greater than s _i is set to 0, T _s1 is the slip rate adjustment of the left front wheel Moment, T _s2 is the slip rate adjustment torque of the right front wheel, T _s3 is the slip rate torque of the left rear wheel, and T _s4 is the slip rate torque of the right rear wheel; The second item of the constraint condition of the objective function: the torque and speed characteristics of the motor; Transform this into the mathematical form of the equality constraint in a quadratic programming problem: T _brmax ≤ T _i ≤ T _drmax Among them, T _brmax is the maximum value of the braking torque that the motor can give, T _i is the torque finally distributed to the four motors by the torque distributor, and T _drmax is the minimum value of the driving torque that the motor can give. The third item of the constraint condition of the objective function: tire force; Transform this into the mathematical form of the equality constraint in a quadratic programming problem: Among them, i=1,2,3,4, which are left front wheel, right front wheel, left rear wheel and right rear wheel respectively, r is wheel radius, μ is road surface adhesion coefficient, T _xi is wheel longitudinal moment, F _yi is the wheel lateral force, F _zi is the wheel vertical force; Organize the three constraints to form the constraints of the quadratic programming problem: 10. The moment distribution method under the steering working condition of four-wheel independent drive electric vehicle as claimed in claim 9, is characterized in that: described solving and carrying out optimal moment distribution: The standard formula of quadratic programming is as follows: Where: W _u is the weight matrix of the driving moment matrix u; u is the driving moment matrix, u=[T ₁ ', T ₂ ', T ₃ , T ₄ ] ^T ; ξ is the weight coefficient; W _v is the weight matrix for assigning errors; b is the linear matrix of quadratic programming; A is the weight matrix of quadratic programming; F _x is the driving force of the wheel, M _x is the yaw moment; δ _f is the front wheel angle, r is the radius of the wheel, l _w is the wheel wheel distance, l _f is the distance from the center of gravity of the vehicle to the front axle; u _min is the matrix of the minimum value of the driving torque, and u _max is the matrix of the maximum value of the driving torque. Construct a geometric sequence to approximate equation (17), so that it can be solved by the effective set method, and the solution process is as follows: Write the inequalities in the constraints in matrix form Select the starting point u ₀ satisfying formula (18), use W to denote the effective constraint index set at this point, set u _k (k=0,1,2,...,N-1) along d _k Direction search, where N is the number of searches for the kth time, and d _k is the step size that descends along the direction of the minimum gradient. The two constraint indicators in formula (18) are regarded as equality constraints, and formula (18) is written as: Among them, u _k is the u value of the kth search, and d _i is the search step size; If u _k + d _i is a feasible solution to the problem, then the step size α _k = 1, u _k+1 = u _k + d _i , and the Lagrangian multiplier can be obtained according to the following formula If the multiplier λ≥0, the optimal solution is u _k+1 , if λ<0, the constraint condition corresponding to the minimum λ is removed from the constraint index set W, and the next iteration of u _k is performed; If u _k + d _i is not a feasible solution to the problem, use formula (21) to find the maximum step size α _k that satisfies the feasible condition, and correct the effective set sequence; α _k ＝max{α _k ∈[0,1]:u _min ≤u _k +α _k d _i ≤u _max } (21) Let u _k+1 ＝u _k +α _k d _i , add this iteration point and the current effective constraint into the working set to start the next update, and so on, and finally find the best point in the feasible region.