Multi-mode multi-target coordination control method for distributed drive-by-wire chassis electric vehicle

By employing a multi-mode, multi-objective coordinated control method, the wheel torque distribution and front wheel steering angle are optimized for different operating conditions of distributed drive drive-by-wire chassis electric vehicles. This solves the problems of high energy consumption at low speeds, large tire wear, and insufficient performance at high speeds in traditional methods, thereby improving the vehicle's safety, stability, and handling performance.

CN118977579APending Publication Date: 2024-11-19CHONGQING JIALING QUANYU MANEUVERING VEHICLE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410977960.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

Traditional distributed drive intelligent electric vehicle control methods increase tire wear and motor power consumption at low speeds, have insufficient operational performance under normal high-speed conditions, and lack safety and stability under extreme high-speed conditions.

Method used

A multi-mode, multi-objective coordinated control method is adopted. Through pattern recognition and state-space equations, an anti-saturation sliding surface and a direct yaw moment algorithm are designed to optimize wheel torque distribution. Combined with linear time-varying model predictive control, coordinated control of front wheel angle and motor torque is achieved.

Benefits of technology

It reduces energy consumption and tire wear at low speeds, improves operational and responsive performance under normal high-speed conditions, and enhances safety and stability under extreme high-speed conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118977579A_ABST
    Figure CN118977579A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-mode multi-target coordination control method for a distributed drive-by-wire chassis electric vehicle, which realizes multi-mode multi-target coordination control, reduces the energy consumption and tire wear of the vehicle at a low speed, improves the operation performance and response performance of the vehicle under a high-speed conventional working condition, and improves the safety of the vehicle. And the safety and stability of the vehicle under the high-speed limit working condition are improved. Comprising the following steps: step 1, based on a current vehicle state and an observed road adhesion coefficient, carrying out classification and identification on a vehicle operation mode in a low-speed working condition, a high-speed conventional working condition and a high-speed limiting working condition based on a rule; 2, low-speed working condition multi-target coordination control is carried out; 3, high-speed conventional working condition multi-target coordination control is carried out; 4, high-speed limit working condition multi-target coordination control is carried out; and 5, controlling the transition process of each working condition.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of distributed drive-by-wire chassis electric vehicles, and in particular to a multi-mode and multi-objective coordinated control method for distributed drive-by-wire chassis electric vehicles. Background Art

[0002] The traditional distributed drive intelligent electric vehicle control method directly analyzes the steering wheel angle and throttle input of the driver into the target yaw rate and longitudinal force requirements, so as to achieve coordinated control of the target yaw rate and longitudinal force through torque vector control. This method has high computational efficiency and mature architecture. However, this method also has corresponding disadvantages. For example, at low speed, the tracking of the reference yaw rate is more suitable to be implemented through the steering wheel angle rather than torque vector control, because compared with high-speed conditions, the same yaw rate at low speed will produce a larger tire slip angle, resulting in torque vector control requiring a larger yaw torque to overcome lateral resistance, increasing tire wear and motor power consumption; while under medium and high speed conventional conditions, the lateral resistance is greatly reduced. At this time, torque vector control is required to quickly and accurately track the driver's required yaw rate and longitudinal torque requirements. Therefore, under this condition, the traditional torque vector control architecture can be used; under medium and high speed extreme conditions, the traditional torque vector control architecture does not fully utilize the redundant control input of the front wheel angle under wire control, and cannot bring the vehicle's performance to the limit, thereby fully ensuring the vehicle's driving safety. Therefore, a multi-mode and multi-objective control method for a distributed drive-by-wire chassis electric vehicle is proposed. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a multi-mode and multi-objective coordinated control method for a distributed drive-by-wire chassis electric vehicle, thereby realizing multi-mode and multi-objective coordinated control, reducing the energy consumption and tire wear of the vehicle at low speeds, improving the operating performance and response performance of the vehicle under high-speed conventional conditions, and improving the safety and stability of the vehicle under high-speed extreme conditions.

[0004] The object of the present invention is achieved in that:

[0005] A multi-mode and multi-objective coordinated control method for a distributed drive-by-wire chassis electric vehicle comprises the following steps:

[0006] Step 1: Pattern recognition of operating conditions

[0007] Based on the current vehicle status and the observed road adhesion coefficient, the vehicle operation mode is classified and identified into low-speed conditions, high-speed normal conditions, and high-speed extreme conditions based on rules.

[0008] Step 2: Multi-objective coordinated control under low-speed conditions

[0009] Compared with high-speed conditions, the same yaw rate will produce a larger tire slip angle under low-speed conditions, resulting in torque vector control requiring a larger yaw moment to overcome lateral resistance, increasing tire wear and motor power consumption; therefore, under low-speed conditions, the reference yaw rate tracking is more suitable for implementation through the active front wheel angle, establishing a two-degree-of-freedom single-track vehicle model and expressing it in the form of a state space equation, x = [β γ] T :

[0010]

[0011] Where k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the yaw rate of the vehicle; δ is the front wheel turning angle of the vehicle; m is the vehicle mass; I z is the moment of inertia in the vertical direction around the center of mass;

[0012] In steady state, The reference yaw rate is obtained as follows:

[0013]

[0014] in,

[0015]

[0016] Therefore, let the difference between the actual and expected yaw rate be e, that is, e = γ - γ ref

[0017] The designed anti-saturation sliding surface S is:

[0018] S=e+λσ (4)

[0019] In the formula, let:

[0020]

[0021] Where θ>0 and is a gain parameter, sat is a saturation function. Based on formulas (4) and (5), the additional front wheel steering angle is obtained as:

[0022]

[0023] where k p =δ max -δ driver sgn(δ driver ), where δ max is the absolute value of the maximum front wheel turning angle, δ driver is the front wheel steering angle obtained by analyzing the driver's steering wheel input. Therefore, the actual front wheel steering angle is the front wheel steering angle obtained by analyzing the driver's steering wheel input plus the additional front wheel steering angle, that is:

[0024]

[0025] The driving / braking force performed by each independent electric drive wheel is analyzed by the driver's accelerator pedal and evenly distributed to obtain:

[0026]

[0027] Step 3: Multi-objective coordinated control of high-speed conventional conditions

[0028] From the perspective of vehicle longitudinal dynamics control requirements, wheel drive / braking force distribution should first meet the driver's longitudinal drive requirements for the vehicle, namely:

[0029] F x1 +F x2 +F x3 +F x4 =ma xd =T tot (9)

[0030] Among them, F xi is the driving / braking force of each independent electric drive wheel; m is the vehicle mass; a xd The longitudinal acceleration of the vehicle required by the driver;

[0031] Electric vehicles use a direct yaw moment algorithm to improve the lateral dynamics of the vehicle itself so that it meets the predetermined motion requirements, that is, to achieve the established lateral dynamics control target of the electric vehicle, where the driving / braking force performed by each independent electric drive wheel needs to generate enough direct yaw moment of the whole vehicle to meet the control requirements, that is:

[0032] l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (10)

[0033] In the formula, l d Half the wheelbase;

[0034] The following uses a direct yaw moment algorithm to improve the lateral dynamic characteristics of the vehicle. First, the reference yaw rate and sideslip angle of the vehicle are calculated based on a two-degree-of-freedom single-track vehicle model.

[0035]

[0036] in:

[0037]

[0038] Among them, k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the vehicle yaw rate; δ is the front wheel turning angle of the vehicle; m is the vehicle mass;

[0039] By considering the equilibrium equation of the two-degree-of-freedom single-track vehicle model and considering the existence of the direct yaw moment Mz:

[0040]

[0041] Expressing the above equation in the form of a state space equation, x = [β γ] T :

[0042]

[0043] in,

[0044]

[0045] Among them, Y β , Y r 、N β 、N r for:

[0046] Y β =-(k1+k2) (17)

[0047]

[0048] N β =-(ak1-bk2) (19)

[0049]

[0050] The required yaw moment is calculated by designing a linear quadratic regulator (LQR).

[0051]

[0052] Among them, Q and R are weighted coefficient matrices,

[0053]

[0054] in:

[0055]

[0056] Then the optimal direct yaw moment is;

[0057]

[0058] After obtaining the direct yaw moment, the calculated direct yaw moment is optimally distributed to the four wheels under constraints. First, an index characterizing the degree of utilization of tire force on road adhesion, namely, the vehicle-road adhesion load index, is established.

[0059]

[0060] where F xi is the longitudinal force of each wheel; F yi is the lateral force of each wheel; F zi is the vertical load of each wheel; μ max is the road adhesion coefficient. Here, it is assumed that the four wheels are on the road with the same road adhesion coefficient. The road adhesion load is simplified and the optimal target is designed:

[0061]

[0062] At the same time, the optimal distribution meets the direct yaw moment and longitudinal force requirements, while considering the vehicle adhesion constraints and motor output limitations:

[0063] F x1 +F x2 +F x3 +F x4 =ma xd (26)

[0064] l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (27)

[0065]

[0066] Among them, F max is the maximum driving / braking force of the motor, where it is assumed that the four motors have the same driving / braking capabilities; d By converting the above problem into a weighted least squares problem, the driving / braking force F of each motor can be quickly solved. x1 , F x2 , F x3 , F x4 ,

[0067] Finally, the driving / braking force performed by each independent electric drive wheel is as follows:

[0068]

[0069] Finally, the required torque of the four motors is obtained, where T fl , Trl 、T fr 、T rr are the torques of the left front wheel, left rear wheel, right front wheel, and right rear wheel respectively, and r is the rolling radius of the tire;

[0070] Step 4: Multi-objective coordinated control of high-speed extreme conditions

[0071] The lateral force is locally linearized at the current operating point. This local linearization is based on the first-order Taylor expansion of the brush tire model around the current sideslip angle. Therefore, the lateral force near the operating point is calculated as follows:

[0072]

[0073] Among them, F yf With F yr Represent the lateral forces on the front and rear axles respectively, is the current lateral force on the tire, is the local cornering stiffness, and the tire force equation is as follows:

[0074]

[0075] where μ max is the road adhesion coefficient, F z is the vertical force, C is the tire cornering stiffness, α is the tire slip angle, σ y =tanα, segmentation point α sl =tan -1 (1 / θ y );

[0076] Secondly, the vehicle lateral dynamics equation is established based on the current working point, and u=[δ f_add ,F x1 ,F x2 ,F x3 ,F x4 ] T , where u is the control input in the lateral dynamics model, including the additional steering wheel angle δ f_add and the longitudinal forces of the four wheels, so the actual front wheel turning angle is the driver’s steering wheel angle δ f_river and the additional steering wheel angle δ output by the controller f_add The superposition is as follows:

[0077] δ=δ f_river +δ f_add (33)

[0078] Therefore, the linear time-varying single-track model can be obtained as follows:

[0079]

[0080] Mz has equality constraints:

[0081] l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (36)

[0082] Among them l d Half the wheelbase;

[0083] Formulas (34) and (35) can be written as state space expressions as follows:

[0084]

[0085] Where:

[0086]

[0087] Then, based on the linearized vehicle lateral dynamics model, a linear time-varying model predictive control problem is constructed to calculate the additional front wheel steering angle and the longitudinal force of each wheel. The calculation method is as follows:

[0088] The reference yaw rate in steady state is calculated based on the driver's steering wheel angle. The reference yaw rate is obtained as follows:

[0089]

[0090] in:

[0091] In order to design the controller, it is necessary to determine the objective function J of LTV-MPC and the constraints. The constraints are divided into hard constraints and soft constraints. Hard constraints strictly restrict the corresponding state variables, inputs or outputs of the system, while soft constraints can be relaxed to avoid the situation where the system has no solution.

[0092] Assuming the controller calculation step size is Ts, and assuming that the driver's steering wheel angle remains constant in each calculation step, the LTV-MPC problem can be described by the following formula:

[0093]

[0094]

[0095] Formula (39) is the objective function, ε is the relaxation factor to ensure that the predictive control problem is solvable, and the first equation in Formula (40) is the equality constraint formed by the Euler discretization of the vehicle lateral dynamics equation, u t+i,tIt is the optimal input quantity at time t+i calculated based on the prediction model at time t. is the reference trajectory at time t+i, assuming that the reference yaw rate It does not change in the prediction time domain, H C and H p They are the system state prediction time domain and the control input prediction time domain, is the weight matrix of the reference trajectory and the actual trajectory, is the weight matrix of the control quantity, ρ i ∈R, in addition, is the side slip angle corresponding to the peak tire force on the rear axle when the road adhesion coefficient is 1, u min =(u min,1 ,u min,2 ,u min,3 ,u min,4 ,u min,5 ,u min,6 ) T , where u min,1 =-δ max ,u min,2 =-min(μ max F z1 ,F motor,1 ),u min,3 =-min(μ max F z2 ,F motor,2 ),u min,4 =-min(μ max F z3 ,F motor,3 ),u min,5 =-min(μ max F z4 ,F motor,4 ),F motor,i (i=1,2,3,4) is the road driving and braking force converted from the peak torque of the motor at the corresponding speed, F zi (i=1,2,3,4) is the vertical load corresponding to each wheel, u min,6 =ma xd , a xd is the longitudinal acceleration required by the driver, which is assumed to be constant in the prediction time domain, u max =(-u min,1 ,-u min,2 ,-u min,3 ,-u min,4 ,-u min,5 ,u min,6 ) T,The above quadratic programming problem is solved by the solver, and finally, the additional front wheel steering angle and the driving and braking torque of each motor are obtained;

[0096] Step 5: Transition process control of each working condition

[0097] In view of the rapid changes and fluctuations in the front wheel angle and the torque of the four motors that may be caused by the switching process of various working conditions, switching control is performed during the working condition switching process to ensure a smooth transition between various working conditions.

[0098] Furthermore, in step 1,

[0099] First, for low-speed conditions, the selection rules for low-speed conditions are as follows:

[0100] Rule 1: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x ≤10km / h;

[0101] Rule 2: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x ≤20km / h;

[0102] If rule 1 or rule 2 is met, the current operating condition is considered as a low-speed condition;

[0103] Secondly, for high-speed conventional working conditions, the selection rules for high-speed conventional working conditions are as follows:

[0104] Rule 3: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x >10km / h,

[0105] And|a y |≤0.8μ max g and

[0106] Rule 4: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x >20km / h,

[0107] And|a y |≤0.8μ max g and

[0108] If rule 3 or rule 4 is met, the current operating condition is considered as high-speed normal operating condition, where a y is the lateral acceleration of the vehicle, ω is the yaw angular velocity of the vehicle, and g is the gravitational acceleration;

[0109] Finally, for the high-speed limit working condition, the high-speed limit working condition selection rules are as follows:

[0110] Rule 5: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x >10km / h and at least one of the following two conditions is met: |a y |>0.8μ max g.

[0111] Rule 6: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x >20km / h and at least one of the following two conditions is met: |a y |>0.8μ max g.

[0112] If Rule 5 or Rule 6 is met, the current operating condition is identified as a high-speed limit condition.

[0113] Furthermore, the vehicle longitudinal acceleration a required by the driver xd Obtained through analysis of the driver's accelerator pedal signal.

[0114] Furthermore, the target front wheel steering angle is obtained by analyzing the driver's steering wheel input.

[0115] Furthermore, in step 5,

[0116] For the front wheel steering angle, during the working condition switching process, if the front wheel does not change, the control input of the front wheel steering angle will not be processed, otherwise the control input of the front wheel steering angle will be processed by a first-order low-pass filter;

[0117] For each drive motor, during the working condition switching process, if the motor torque does not change, the control input of the motor torque is not processed, otherwise the control input of the motor torque is subjected to a first-order low-pass filter.

[0118] Due to the adoption of the above technical solution, the present invention has the following beneficial effects:

[0119] Through the current vehicle state and the observed road adhesion coefficient, the vehicle's operating mode is classified into low-speed conditions, high-speed conventional conditions, and high-speed extreme conditions based on rules, and corresponding control methods and transition process control methods are designed; multi-mode and multi-objective coordinated control is achieved, which reduces the vehicle's energy consumption and tire wear at low speeds, improves the vehicle's operating performance and response performance under high-speed conventional conditions, and improves the vehicle's safety and stability under high-speed extreme conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] Figure 1 It is a schematic diagram of the overall structure of the present invention;

[0121] Figure 2 A schematic diagram is selected for the mode of the present invention;

[0122] Figure 3 It is a schematic diagram of the control principle of low-speed working condition of the present invention;

[0123] Figure 4 This is a schematic diagram of the control principle of the high-speed conventional working condition of the present invention;

[0124] Figure 5 It is a schematic diagram of the transition process control principle of the present invention. DETAILED DESCRIPTION

[0125] The technical problem to be solved by the present invention is mainly applied to distributed drive-by-wire chassis electric vehicles, and a multi-mode multi-objective coordinated control method suitable for low-speed conditions, high-speed conventional conditions and high-speed extreme conditions is designed. The main solution is the problem that the multi-objective coordinated control method in a single mode cannot take into account the low energy consumption requirements and less tire wear at low speeds, the precise and rapid tracking of the driver's maneuverability requirements under high-speed conventional conditions, and the safe and stable vehicle control under high-speed extreme conditions.

[0126] A multi-mode and multi-objective coordinated control method for a distributed drive-by-wire chassis electric vehicle comprises the following steps:

[0127] Step 1: Pattern recognition of operating conditions

[0128] Mode recognition of operating conditions refers to the classification and identification of vehicle operating modes into low-speed conditions, high-speed normal conditions, and high-speed extreme conditions based on the current vehicle state (longitudinal speed, yaw angular velocity, longitudinal / lateral acceleration, etc.) and the observed road adhesion coefficient.

[0129] The details are as follows:

[0130] First, for low-speed conditions, the selection rules for low-speed conditions are as follows:

[0131] Rule 1: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x ≤10km / h;

[0132] Rule 2: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x ≤20km / h;

[0133] If Rule 1 or Rule 2 is met, the current operating condition is considered a low-speed condition. It must be emphasized that the thresholds set for the road adhesion coefficient and longitudinal vehicle speed in Rules 1 and 2 are allowed to be appropriately adjusted in actual tests.

[0134] Secondly, for high-speed conventional working conditions, the selection rules are as follows:

[0135] Rule 3:

[0136] Road adhesion coefficient μ max ≤0.5 and longitudinal speed v x >10km / h

[0137] And|a y |≤0.8μ max g and

[0138] Rule 4:

[0139] Road adhesion coefficient 0.5<μ max And the longitudinal speed v x >20km / h

[0140] And|a y |≤0.8μ max g and

[0141] If rule 3 or rule 4 is met, the current operating condition is considered as high-speed normal operating condition, where a y is the lateral acceleration of the vehicle, ω is the yaw angular velocity of the vehicle, and g is the acceleration of gravity. It must be emphasized that the thresholds set in Rules 3 and 4 can be appropriately adjusted in actual tests.

[0142] Finally, for high-speed extreme conditions, the selection rules are as follows:

[0143] Rule 5:

[0144] Road adhesion coefficient μ max ≤0.5 and longitudinal speed v x >10km / h

[0145] And (|a y |>0.8μ max g or

[0146] Rule 6:

[0147] Road adhesion coefficient 0.5<μ max And the longitudinal speed v x >20km / h

[0148] And (|a y |>0.8μ max g or

[0149] If Rule 5 or Rule 6 is met, the current operating condition is identified as a high-speed limit condition. It must be emphasized that for Rule 5, under the premise of meeting the road adhesion coefficient and longitudinal vehicle speed conditions, only one of the lateral acceleration and yaw rate needs to be met to meet Rule 5; similarly, for Rule 6, under the premise of meeting the road adhesion coefficient and longitudinal vehicle speed conditions, only one of the lateral acceleration and yaw rate needs to be met to meet Rule 6. In addition, the thresholds set in Rules 5 and 6 are allowed to be appropriately adjusted in actual tests.

[0150] Step 2: Multi-objective coordinated control under low-speed conditions

[0151] Compared with high-speed conditions, the same yaw rate at low speed will produce a larger tire slip angle, resulting in torque vector control requiring a larger yaw moment to overcome lateral resistance, increasing tire wear and motor power consumption; therefore, at low speeds, reference yaw rate tracking is more suitable to be implemented through active front wheel steering.

[0152] In order to better describe the motion state of the vehicle under stable working conditions, a two-degree-of-freedom single-track vehicle model is established and expressed in the form of a state space equation, x = [β γ] T :

[0153]

[0154] Where k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the yaw rate of the vehicle; δ is the front wheel turning angle of the vehicle; m is the mass of the vehicle; I z is the moment of inertia in the vertical direction about the center of mass.

[0155] In steady state, The reference yaw angular velocity is obtained as follows:

[0156]

[0157] in

[0158]

[0159] Therefore, let the difference between the actual and expected yaw rate be e, that is, e = γ - γ ref

[0160] The anti-saturation sliding surface S is designed as

[0161] S=e+λσ (4)

[0162] In the formula, and another

[0163]

[0164] Where θ>0 and is a gain parameter, sat is a saturation function. Based on formulas (4) and (5), the additional front wheel steering angle is

[0165]

[0166] Among them, k p =δ max -δ driver sgn(δ driver )

[0167] Among them, δ max is the absolute value of the maximum front wheel turning angle, δ driver is the front wheel steering angle obtained by the driver’s steering wheel input. Therefore, the actual front wheel steering angle is the front wheel steering angle obtained by the driver’s steering wheel input plus the additional front wheel steering angle, that is,

[0168]

[0169] The driving / braking force performed by each independent electric drive wheel is analyzed by the driver's accelerator pedal and evenly distributed to obtain:

[0170]

[0171] Step 3: Multi-objective coordinated control of high-speed conventional conditions

[0172] From the perspective of vehicle longitudinal dynamics control requirements, wheel drive / braking force distribution should first meet the driver's longitudinal drive requirements for the vehicle, that is;

[0173] F x1 +F x2 +F x3 +F x4 =ma xd =T tot (9)

[0174] where F xi is the driving / braking force of each independent electric drive wheel; m is the vehicle mass; a xd The longitudinal acceleration of the vehicle required by the driver can be obtained by analyzing the driver's accelerator pedal signal.

[0175] On the other hand, when the electric vehicle direct yaw moment algorithm is used to improve the lateral dynamic characteristics of the vehicle itself so that it meets the predetermined motion requirements, that is, to achieve the established electric vehicle lateral dynamic control target, the driving / braking force performed by each independent electric drive wheel needs to generate sufficient direct yaw moment of the whole vehicle to meet the control requirements, that is:

[0176] l d (F x2 -Fx1 )+l d (F x4 -F x3 )=Mz (10)

[0177] Among them l d Half the wheelbase.

[0178] The direct yaw moment algorithm is used below to improve the lateral dynamic characteristics of the vehicle. Firstly, the reference yaw rate and sideslip angle of the vehicle are calculated based on the two-degree-of-freedom single-track vehicle model.

[0179]

[0180] in:

[0181]

[0182] Where k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the yaw rate of the vehicle; δ is the front wheel turning angle of the vehicle; and m is the mass of the vehicle.

[0183] By considering the equilibrium equation of the two-degree-of-freedom single-track vehicle model and considering the existence of the direct yaw moment Mz

[0184]

[0185] Expressing the above equation in the form of a state space equation, x = [β γ] T :

[0186]

[0187] in

[0188]

[0189] Among them, Y β , Y r 、N β 、N r for

[0190] Y β =-(k1+k2) (17)

[0191]

[0192] N β =-(ak1-bk2) (19)

[0193]

[0194] The required yaw moment is calculated by designing a linear quadratic regulator (LQR).

[0195]

[0196] Where Q and R are weighted coefficient matrices,

[0197]

[0198] in:

[0199]

[0200] Then the optimal direct yaw moment is;

[0201]

[0202] After obtaining the direct yaw moment, the calculated direct yaw moment is optimally distributed to the four wheels under certain constraints. First, an index characterizing the degree of utilization of tire force on road adhesion is established, namely, the vehicle-road adhesion load index.

[0203]

[0204] where F xi is the longitudinal force of each wheel; F yi is the lateral force of each wheel; F zi is the vertical load of each wheel; μ max is the road adhesion coefficient. Here it is assumed that the four wheels are on the road with the same road adhesion coefficient.

[0205] Simplify the road adhesion load and design the optimal target

[0206]

[0207] At the same time, the optimal distribution meets the direct yaw moment and longitudinal force requirements, while considering the vehicle adhesion constraints and motor output limitations:

[0208] F x1 +F x2 +F x3 +F x4 =ma xd (26)

[0209] l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (27)

[0210]

[0211] where F max is the maximum driving / braking force of the motor, where it is assumed that the four motors have the same driving / braking capabilities; d By converting the above problem into a weighted least squares problem, the driving / braking force F of each motor can be quickly solved. x1 , F x2 , F x3 , F x4 .

[0212] Finally, the driving / braking force performed by each independent electric drive wheel is as follows:

[0213]

[0214] Finally, the required torque of the four motors can be obtained, where T fl 、T rl 、T fr 、T rr are the torques of the left front wheel, left rear wheel, right front wheel, and right rear wheel respectively; r is the rolling radius of the tire. The target turning angle of the front wheel is obtained by analyzing the input of the driver's steering wheel.

[0215] Step 4: Multi-objective coordinated control of high-speed extreme conditions

[0216] Under medium and high speed extreme conditions, the traditional torque vector control architecture does not fully utilize the redundant control input of the front wheel angle under wire control, and cannot bring the vehicle's performance to the limit, thereby fully ensuring the vehicle's driving safety; at the same time, because in the linear single-track model, the cornering stiffness is selected as the approximate slope of the linear segment in the tire cornering characteristic curve, which is constant and can only accurately reflect the dynamic response of the vehicle in the linear stage, in order to ensure the control effect of the model predictive control, it is necessary to ensure the matching of the model in both the linear domain and the nonlinear domain. In this paper, the additional front wheel angle is applied, and the nonlinear vehicle model is linearized near the working point. These elements are integrated into the optimization problem of multi-objective coordinated control under high-speed extreme conditions to achieve safe and stable control of the vehicle under high-speed extreme conditions.

[0217] First, we introduce the local linearization of the lateral force at the current operating point. This local linearization is based on the first-order Taylor expansion of the brush tire model around the current slip angle. Therefore, the lateral force near the operating point can be calculated as follows:

[0218]

[0219] Among them, F yf With F yr Represent the lateral forces on the front and rear axles respectively. is the current lateral force on the tire, is the local cornering stiffness, and the tire force equation is as follows:

[0220]

[0221] where μ max is the road adhesion coefficient, F z is the vertical force, C is the tire cornering stiffness, α is the tire slip angle, σ y =tanα, segmentation point α sl =tan -1 (1 / θ y ).

[0222] Secondly, we establish the vehicle lateral dynamics equation based on the current working point. Let u = [δ f_add ,F x1 ,F x2 ,F x3 ,F x4 ] T , where u is the control input in the lateral dynamics model we study, including the additional steering wheel angle δ f_add and the longitudinal forces of the four wheels. Therefore, the actual front wheel turning angle is the driver's steering wheel angle δ f_river and the additional steering wheel angle δ output by the controller f_add The superposition is as follows:

[0223] δ=δ f_river +δ f_add (33)

[0224] Therefore, the linear time-varying single-track model can be obtained as follows:

[0225]

[0226] Mz exists equality constraint

[0227] l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (36)

[0228] Among them l d Half the wheelbase.

[0229] Formulas (34) and (35) can be written as state space expressions as follows:

[0230]

[0231] Where:

[0232]

[0233] Then, we construct a linear time-varying model predictive control problem (LTV-MPC) based on the linearized vehicle lateral dynamics model to calculate the additional front wheel steering angle and the longitudinal force of each wheel as follows:

[0234] The reference yaw rate in steady state is calculated based on the driver's steering wheel angle. The reference yaw rate can be obtained from this:

[0235]

[0236] in:

[0237] In order to design the controller, it is necessary to determine the objective function J and constraints of LTV-MPC. The constraints are divided into hard constraints and soft constraints. Hard constraints strictly restrict the corresponding state variables, inputs or outputs of the system, while soft constraints can be relaxed to avoid the situation where the system has no solution.

[0238] Assume that the controller calculation step is Ts, and assume that the driver's steering wheel angle remains constant in each calculation step. Then the LTV-MPC problem can be described by the following formula:

[0239]

[0240]

[0241] Formula (39) is the objective function, and ε is the relaxation factor to ensure that the predictive control problem is solvable. The first equation in formula (40) is the equality constraint formed by the Euler discretization of the vehicle lateral dynamics equation. t+i,t It is the optimal input quantity at time t+i calculated based on the prediction model at time t. is the reference trajectory at time t+i, where we assume that the reference yaw rate Does not change during the prediction time domain. C and H p They are the system state prediction domain and the control input prediction domain respectively. is the weight matrix of the reference trajectory and the actual trajectory, is the weight matrix of the control quantity, ρ i ∈R. In addition, α max is the sideslip angle corresponding to the peak tire force on the rear axle when the road adhesion coefficient is 1. u min =(u min,1,u min,2 ,u min,3 ,u min,4 ,u min,5 ,u min,6 ) T , where u min,1 =-δ max ,u min,2 =-min(μ max F z1 ,F motor,1 ),u min,3 =- min (μ max F z2 ,F motor,2 ),u min,4 =-min(μ max F3,F motor,3 ),u min,5 =-min(μ max F z4 ,F motor,4 ),F motor,i (i=1,2,3,4) is the road driving and braking force converted from the peak torque of the motor at the corresponding speed, F zi (i=1,2,3,4) is the vertical load corresponding to each wheel, u min,6 =ma xd , a xd is the longitudinal acceleration required by the driver, which is assumed to be constant in the prediction time domain, u max =(-u min,1 ,-u min,2 ,-u min,3 ,-u min,4 ,-u min,5 ,u min,6 ) T The above quadratic programming problem can be solved by some solvers, such as OSQP. Finally, the additional front wheel steering angle and the driving and braking torque of each motor are obtained.

[0242] Step 5: Transition process control of each working condition

[0243] In view of the rapid changes and fluctuations in the front wheel angle and the torque of the four motors that may be caused by the switching process of various working conditions, step 5 performs switching control during the working condition switching process to ensure a smooth transition between various working conditions.

[0244] The specific steps are as follows:

[0245] For the front wheel steering angle, during the working condition switching process, if the front wheel does not change, the control input of the front wheel steering angle will not be processed, otherwise the control input of the front wheel steering angle will be processed by a first-order low-pass filter. The processing process lasts for 0.5 seconds. This 0.5-second threshold is allowed to be appropriately adjusted during the actual test process.

[0246] For each drive motor, during the working condition switching process, if the motor torque does not change, the control input of the motor torque will not be processed, otherwise the control input of the motor torque will be processed by a first-order low-pass filter. The processing process lasts for 0.5 seconds. This 0.5-second threshold is allowed to be appropriately adjusted during the actual test process.

[0247] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A multi-mode and multi-objective coordinated control method for a distributed drive-by-wire chassis electric vehicle, characterized in that The following steps are involved: Step 1: Pattern recognition of operating conditions Based on the current vehicle state and the observed road adhesion coefficient, the vehicle operation mode is classified and identified into low-speed conditions, high-speed normal conditions, and high-speed extreme conditions based on rules; Step 2: Multi-objective coordinated control under low-speed conditions Compared with high-speed conditions, the same yaw rate will produce a larger tire slip angle under low-speed conditions, resulting in torque vector control requiring a larger yaw moment to overcome lateral resistance, increasing tire wear and motor power consumption; therefore, under low-speed conditions, the reference yaw rate tracking is more suitable for implementation through the active front wheel angle, establishing a two-degree-of-freedom single-track vehicle model and expressing it in the form of a state space equation, x = [β γ] T : Where k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the yaw rate of the vehicle; δ is the front wheel turning angle of the vehicle; m is the vehicle mass; I z is the moment of inertia in the vertical direction around the center of mass; In steady state, The reference yaw rate is obtained as follows: in, Therefore, let the difference between the actual and expected yaw rate be e, that is, e = γ - γ ref The designed anti-saturation sliding surface S is: S=e+λσ (4) In the formula, let: Where θ>0 and is a gain parameter, sat is a saturation function. Based on formulas (4) and (5), the additional front wheel steering angle is obtained as: Among them, k p =δ max -δ driver sgn(δ driver ), where δ max is the absolute value of the maximum front wheel turning angle, δ driver is the front wheel steering angle obtained by analyzing the driver's steering wheel input. Therefore, the actual front wheel steering angle is the front wheel steering angle obtained by analyzing the driver's steering wheel input plus the additional front wheel steering angle, that is: The driving / braking force performed by each independent electric drive wheel is analyzed by the driver's accelerator pedal and evenly distributed to obtain: Step 3: Multi-objective coordinated control of high-speed conventional conditions From the perspective of vehicle longitudinal dynamics control requirements, wheel drive / braking force distribution should first meet the driver's longitudinal drive requirements for the vehicle, namely: F x1 +F x2 +F x3 +F x4 =at xd =T tot (9) Among them, F xi is the driving / braking force of each independent electric drive wheel; m is the vehicle mass; a xd The longitudinal acceleration of the vehicle required by the driver; Electric vehicles use a direct yaw moment algorithm to improve the lateral dynamics of the vehicle itself so that it meets the predetermined motion requirements, that is, to achieve the established lateral dynamics control target of the electric vehicle, where the driving / braking force performed by each independent electric drive wheel needs to generate enough direct yaw moment of the whole vehicle to meet the control requirements, that is: l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (10) In the formula, l d Half the wheelbase; The following uses a direct yaw moment algorithm to improve the lateral dynamic characteristics of the vehicle. First, the reference yaw rate and sideslip angle of the vehicle are calculated based on a two-degree-of-freedom single-track vehicle model. in: Among them, k1 and k2 are the cornering stiffness of the front and rear axles; a and b are the distances from the center of mass to the front and rear axles; β is the side slip angle of the vehicle center of mass; γ is the vehicle yaw rate; δ is the front wheel turning angle of the vehicle; m is the vehicle mass; By considering the equilibrium equation of the two-degree-of-freedom single-track vehicle model and considering the existence of the direct yaw moment Mz: Expressing the above equation in the form of a state space equation, x = [β γ] T : in, Among them, Y β , Y r 、N β 、N r for: AND β =-(k1+k2) (17) N β =-(ak1-bk2) (19) The required yaw moment is calculated by designing a linear quadratic regulator (LQR). Among them, Q and R are weighted coefficient matrices, in: Then the optimal direct yaw moment is; After obtaining the direct yaw moment, the calculated direct yaw moment is optimally distributed to the four wheels under constraints. First, an index characterizing the degree of utilization of tire force on road adhesion, namely, the vehicle-road adhesion load index, is established. where F xi is the longitudinal force of each wheel; F yi is the lateral force of each wheel; F zi is the vertical load of each wheel; μ max is the road adhesion coefficient. Here, it is assumed that the four wheels are on the road with the same road adhesion coefficient. The road adhesion load is simplified and the optimal target is designed: At the same time, the optimal distribution meets the direct yaw moment and longitudinal force requirements, while considering the vehicle adhesion constraints and motor output limitations: F x1 +F x2 +F x3 +F x4 =at xd (26) l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (27) Among them, F max is the maximum driving / braking force of the motor, where it is assumed that the four motors have the same driving / braking capabilities; d By converting the above problem into a weighted least squares problem, the driving / braking force F of each motor can be quickly solved. x1 , F x2 , F x3 , F x4 , Finally, the driving / braking force performed by each independent electric drive wheel is as follows: Finally, the required torque of the four motors is obtained, where T fl , T rl , T fr , T rr are the torques of the left front wheel, left rear wheel, right front wheel, and right rear wheel respectively, and r is the rolling radius of the tire; Step 4: Multi-objective coordinated control of high-speed extreme conditions The lateral force is locally linearized at the current operating point. This local linearization is based on the first-order Taylor expansion of the brush tire model around the current sideslip angle. Therefore, the lateral force near the operating point is calculated as follows: Among them, F yf With F yr Represent the lateral forces on the front and rear axles respectively, is the current lateral force on the tire, is the local cornering stiffness, and the tire force equation is as follows: where μ max is the road adhesion coefficient, F z is the vertical force, C is the tire cornering stiffness, α is the tire slip angle, σ y =tanα, segmentation point α sl =tan -1 (1 / θ y ); Secondly, the vehicle lateral dynamics equation is established based on the current working point, and u=[δ f_add ,F x1 ,F x2 ,F x3 ,F x4 ] T , where u is the control input in the lateral dynamics model, including the additional steering wheel angle δ f_add and the longitudinal forces of the four wheels, so the actual front wheel turning angle is the driver’s steering wheel angle δ f_river and the additional steering wheel angle δ output by the controller f_add The superposition is shown as follows: d=d f_river +d f_add (33) Therefore, the linear time-varying single-track model can be obtained as follows: Mz has equality constraints: l d (F x2 -F x1 )+l d (F x4 -F x3 )=Mz (36) Among them l d Half the wheelbase; Formulas (34) and (35) can be written as state space expressions as follows: Where: Then, based on the linearized vehicle lateral dynamics model, a linear time-varying model predictive control problem is constructed to calculate the additional front wheel steering angle and the longitudinal force of each wheel. The calculation method is as follows: The reference yaw rate in steady state is calculated based on the driver's steering wheel angle. The reference yaw rate is obtained as follows: in: In order to design the controller, it is necessary to determine the objective function J of LTV-MPC and the constraints. The constraints are divided into hard constraints and soft constraints. Hard constraints strictly restrict the corresponding state variables, inputs or outputs of the system, while soft constraints can be relaxed to avoid the situation where the system has no solution. Assuming the controller calculation step size is Ts, and assuming that the driver's steering wheel angle remains constant in each calculation step, the LTV-MPC problem can be described by the following formula: Formula (39) is the objective function, ε is the relaxation factor to ensure that the predictive control problem is solvable, and the first equation in Formula (40) is the equality constraint formed by the Euler discretization of the vehicle lateral dynamics equation, u t+i,t It is the optimal input quantity at time t+i calculated based on the prediction model at time t. is the reference trajectory at time t+i, assuming that the reference yaw rate It does not change in the prediction time domain, H c and H p They are the system state prediction time domain and the control input prediction time domain, is the weight matrix of the reference trajectory and the actual trajectory, is the weight matrix of the control quantity, ρ i ∈R, in addition, α max is the side slip angle corresponding to the peak tire force on the rear axle when the road adhesion coefficient is 1, u min =(u min,1 ,u min,2 ,u min,3 ,u min,4 ,u min,5 ,u min,6 ) T , where u min,1 =-δ max ,u min,2 =-min(μ max F z1 ,F motor,1 ),u min,3 =-min(μ max F z2 ,F motor,2 ),u min,4 =-min(μ max F z3 ,F motor,3 ),u min,5 =-min(μ max F z4 ,F motor,4 ),F motor,I (=1,2,3,4) is the road driving and braking force converted from the peak torque of the motor at the corresponding speed, F zi (i=1,2,3,4) is the vertical load corresponding to each wheel, u min,6 =ma xd , a xd is the longitudinal acceleration required by the driver, which is assumed to be constant in the prediction time domain, u max =(-u min,1 ,-u min,2 ,-u min,3 ,-u min,4 ,-u min,5 ,u min,6 ) T ,The above quadratic programming problem is solved by the solver, and finally, the additional front wheel steering angle and the driving and braking torque of each motor are obtained; Step 5: Transition process control of each working condition In view of the rapid changes and fluctuations in the front wheel angle and the torque of the four motors that may be caused by the switching process of various working conditions, switching control is performed during the working condition switching process to ensure a smooth transition between various working conditions.

2. The multi-mode and multi-objective coordinated control method of a distributed drive-by-wire chassis electric vehicle according to claim 1, characterized in that: In step 1, First, for low-speed conditions, the selection rules for low-speed conditions are as follows: Rule 1: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x ≤10km / h; Rule 2: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x ≤20km / h; If rule 1 or rule 2 is met, the current operating condition is considered as a low-speed condition; Secondly, for high-speed conventional working conditions, the selection rules for high-speed conventional working conditions are as follows: Rule 3: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x >10km / h, And|a y |≤0.8μ max g and Rule 4: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x >20km / h, And|a y |≤0.8μ amx g and If rule 3 or rule 4 is met, the current operating condition is considered as high-speed normal operating condition, where a y is the lateral acceleration of the vehicle, ω is the yaw angular velocity of the vehicle, and g is the gravitational acceleration; Finally, for the high-speed limit working condition, the high-speed limit working condition selection rules are as follows: Rule 5: Road Adhesion Coefficient μ max ≤0.5 and longitudinal speed v x >10km / h and at least one of the following two conditions is met: |a y |>0.8μ max g. Rule 6: Road Adhesion Coefficient 0.5 < μ max And the longitudinal speed v x >20km / h and at least one of the following two conditions is met: |a y |>0.8μ max g. If Rule 5 or Rule 6 is met, the current operating condition is identified as a high-speed limit condition.

3. The multi-mode and multi-objective coordinated control method of a distributed drive-by-wire chassis electric vehicle according to claim 1, characterized in that: The vehicle longitudinal acceleration a required by the driver xd Obtained through analysis of the driver's accelerator pedal signal.

4. The multi-mode and multi-objective coordinated control method of a distributed drive-by-wire chassis electric vehicle according to claim 1, characterized in that: The target front wheel angle is analyzed from the driver's steering wheel input.

5. The multi-mode and multi-objective coordinated control method of a distributed drive-by-wire chassis electric vehicle according to claim 1, characterized in that: In step 5, For the front wheel steering angle, during the working condition switching process, if the front wheel does not change, the control input of the front wheel steering angle will not be processed, otherwise the control input of the front wheel steering angle will be processed by a first-order low-pass filter; For each drive motor, during the working condition switching process, if the motor torque does not change, the control input of the motor torque is not processed, otherwise the control input of the motor torque is subjected to a first-order low-pass filter.