A vehicle design optimization method and device coordinating optimal design and optimal control

Abstract

The application discloses a vehicle design optimization method and device based on collaborative optimal design and optimal control, and relates to the technical field of vehicle design optimization. In the method, a multi-degree-of-freedom vehicle dynamics model is constructed based on Lagrange dynamics, and a power assembly mass model is constructed. Before constructing a collaborative optimization problem, simulation control is performed under a preset working condition based on the multi-degree-of-freedom vehicle dynamics model and the power assembly mass model to obtain an initial solution. Then, a multi-objective collaborative optimization problem is constructed according to the initial solution, a design parameter set, a control parameter set and related constraints. Finally, the multi-objective collaborative optimization problem is converted into a nonlinear optimization problem, and a solver is used to solve the problem to obtain an optimal control parameter set and an optimal design parameter set. The application constructs a collaborative optimization problem for a vehicle, solves the problem after conversion into a nonlinear optimization problem, and can simultaneously optimize the design parameters and the control parameters of the vehicle.

Description

Technical Field

[0001] This application relates to the field of vehicle design optimization technology, and in particular to a vehicle design optimization method and apparatus that combines optimal design and optimal control. Background Technology

[0002] In recent years, the pace of new car launches has accelerated, the global automotive market has changed rapidly, intelligent vehicles have developed at a breakneck pace, product iterations have been swift, the global vehicle R&D and delivery cycle has shortened dramatically, performance requirements have been continuously increasing, and the elimination of the cockpit in driverless cars has created numerous potential forms that urgently require rapid development and iteration. New R&D processes are needed to shorten the production cycle in platform interconnection and software areas. Currently, the average R&D cycle for a new car has been shortened from 5-7 years to 2-3 years. The design and development of vehicles in the new era presents the following demands: more diverse forms, shorter cycles, and higher requirements.

[0003] In traditional new vehicle development, existing vehicle design methods primarily rely on experience. These methods combine existing research findings to improve the vehicle body and select suitable powertrain components, offering advantages such as short design cycles and high specificity. However, this approach mainly relies on qualitative analysis based on experience and cannot theoretically guarantee that the solution will be optimal in terms of structure and performance. In vehicle lightweight design, traditional methods can only select specific optimization solutions based on qualitative evaluations, resulting in low optimization efficiency and no guarantee of optimal results. Furthermore, they cannot combine specific analysis results to perform targeted parameter optimization for different parts of the vehicle.

[0004] Currently, vehicle parameter optimization methods can be categorized into several types: 1) Static model-based parameter optimization methods, while computationally efficient, suffer from poor practical application due to the dynamic nature of vehicle operation. 2) Evolutionary algorithm-based optimization methods, which consider dynamic vehicle models, but lack gradient information, resulting in low optimization efficiency. 3) Dynamic programming or indirect methods based on dynamic models, which are highly sensitive to the problem's scale and struggle to achieve rapid optimization of numerous parameters.

[0005] Furthermore, the vehicle's design parameters and control parameters are coupled, while the aforementioned optimization methods only optimize either the design or control parameters, significantly reducing the global optimization space and limiting the potential for vehicle performance improvement. Optimizing the design and control parameters separately makes it difficult to achieve optimal performance and can expand the optimization space, but the resulting problem is a multi-degree-of-freedom optimal design and optimal control problem with a large parameter scale, making it extremely challenging to solve, and making co-optimization of design and control parameters difficult. Summary of the Invention

[0006] The purpose of this application is to provide a vehicle design optimization method and apparatus that combines optimal design and optimal control, which can simultaneously optimize the vehicle's design parameters and control parameters based on a vehicle dynamics model.

[0007] To achieve the above objectives, this application provides the following solution:

[0008] Firstly, this application provides a vehicle design optimization method that combines collaborative optimal design and optimal control, comprising the following steps:

[0009] Based on the vehicle type, a multi-degree-of-freedom vehicle dynamics model is constructed using Lagrange dynamics.

[0010] Based on the dependence of the powertrain mass and output torque on design parameters, a powertrain mass model is constructed.

[0011] Based on the multi-degree-of-freedom vehicle dynamics model and powertrain mass model, simulation control is performed under preset working conditions to obtain the initial solution.

[0012] Based on the multi-degree-of-freedom vehicle dynamics model and powertrain mass model, a multi-objective collaborative optimization problem is constructed according to the initial solution, design parameter set, control parameter set and related constraints.

[0013] The multi-objective collaborative optimization problem is transformed into a nonlinear optimization problem, and a nonlinear optimization problem solver is used to solve the problem, obtaining the optimal control parameter set and the optimal design parameter set.

[0014] Secondly, this application provides a vehicle design optimization device for collaborative optimal design and optimal control, comprising the following modules:

[0015] The dynamics model building module is used to construct multi-degree-of-freedom vehicle dynamics models based on Lagrange dynamics, according to the vehicle type.

[0016] The mass model building module is used to construct a powertrain mass model based on the dependence of the powertrain mass and output torque on design parameters.

[0017] The initial solution generation module is used to perform simulation control under preset working conditions based on a multi-degree-of-freedom vehicle dynamics model and a powertrain mass model to obtain the initial solution.

[0018] The multi-objective optimization problem construction module is used to construct multi-objective collaborative optimization problems based on multi-degree-of-freedom vehicle dynamics models and powertrain mass models, according to the initial solution, design parameter set, control parameter set, and relevant constraints.

[0019] The problem transformation and NLP solving module is used to transform multi-objective collaborative optimization problems into nonlinear optimization problems, and to solve the problems using a nonlinear optimization problem solver to obtain the optimal control parameter set and the optimal design parameter set.

[0020] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0021] This application provides a vehicle design optimization method and apparatus for coordinated optimal design and optimal control. The method first constructs a multi-degree-of-freedom vehicle dynamics model based on Lagrange dynamics, and then constructs a powertrain mass model based on the dependence of powertrain mass and output torque on design parameters. Before constructing the coordinated optimization problem, simulation control is performed under preset operating conditions based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model to obtain an initial solution. Then, based on the initial solution, design parameter set, control parameter set, and relevant constraints, a multi-objective coordinated optimization problem is constructed. This multi-objective coordinated optimization problem is then transformed into a nonlinear optimization problem, and solved using a nonlinear optimization problem solver to obtain the optimal control parameter set and the optimal design parameter set. This application constructs a vehicle-oriented coordinated optimal design and optimal control problem, enabling simultaneous optimization of design and control parameters based on the vehicle dynamics model. The optimal design parameters can be used for vehicle design, and the optimal control parameters can serve as a reference for subsequent online control strategy formulation. The scheme of this application can apply gradient information to expand the optimization space and achieves good optimization results. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating a vehicle design optimization method based on collaborative optimal design and optimal control, provided as an embodiment of this application.

[0024] Figure 2 This is a schematic diagram of the vehicle dynamics model architecture in a vehicle design optimization method of collaborative optimal design and optimal control provided in an embodiment of this application.

[0025] Figure 3 This is a schematic diagram of a vehicle with 14 degrees of freedom in a vehicle design optimization method of collaborative optimal design and optimal control provided in an embodiment of this application.

[0026] Figure 4This is a schematic diagram of the vehicle's xy plane in a vehicle design optimization method for collaborative optimal design and optimal control provided in an embodiment of this application.

[0027] Figure 5 This is a schematic diagram of tire-road contact in a vehicle design optimization method of collaborative optimal design and optimal control provided in an embodiment of this application.

[0028] Figure 6 This is a schematic diagram of vehicle forces in a vehicle design optimization method based on collaborative optimal design and optimal control, provided as an embodiment of this application.

[0029] Figure 7 This is a schematic diagram of the forces and moments acting on unsprung mass in a vehicle design optimization method of collaborative optimal design and optimal control provided in an embodiment of this application.

[0030] Figure 8 This is a schematic diagram of the XY coordinates of an example track in a vehicle design optimization method for collaborative optimal design and optimal control provided in an embodiment of this application.

[0031] Figure 9 This is a schematic diagram of the trajectory curvature of an example track in a vehicle design optimization method for collaborative optimal design and optimal control provided in an embodiment of this application.

[0032] Figure 10 This is a schematic diagram of the vehicle's attitude in a curvilinear coordinate system in a vehicle design optimization method for collaborative optimal design and optimal control provided in an embodiment of this application.

[0033] Figure 11 This is a schematic diagram of the collaborative optimal design and optimal control algorithm framework used in a vehicle design optimization method based on a collaborative optimal design and optimal control provided in an embodiment of this application.

[0034] Figure 12 This is a schematic diagram of the functional modules of a vehicle design optimization device for collaborative optimal design and optimal control, provided in an embodiment of this application. Detailed Implementation

[0035] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0036] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] In one exemplary embodiment, such as Figure 1 As shown, a vehicle design optimization method based on collaborative optimal design and optimal control is provided, including the following steps:

[0038] A1. Based on the vehicle type, a multi-degree-of-freedom vehicle dynamics model is constructed using Lagrange dynamics. The multi-degree-of-freedom vehicle dynamics model can be expressed by the following equation:

[0039]

[0040] Where T is the kinetic energy of the vehicle's power system, and Q is... b Q is the generalized force acting on the sprung mass. u For the generalized force acting on the unsprung mass, q b q represents the degrees of freedom of the vehicle body. u X represents the rotation and vertical displacement of the four tires. A Y A and Z A These represent the absolute displacements of the centroid along the geodetic coordinate system. φ and ψ represent the rotation angles of the vehicle body about the three axes of the geodetic coordinate system, respectively; θ ufr ,θ ufl ,θ urr ,θ url These represent the angular velocity sequences generated by the rotational motions of the four tires: right front, left front, right rear, and left rear. ufr ,z ufl ,z urr ,z url These represent the vertical displacement sequences of the four tires: right front, left front, right rear, and left rear.

[0041] The kinetic energy of a vehicle's powertrain includes the kinetic energy of sprung mass and the kinetic energy of unsprung mass; the kinetic energy of sprung mass can be expressed by the following formula:

[0042]

[0043] Among them, T b V is the kinetic energy of the sprung mass. b M represents the component of the generalized velocity of the vehicle body in the vehicle body coordinate system. b Let V be the mass matrix of the vehicle body. u,i Let M be the velocity of the i-th tire in the vehicle coordinate system. u,i Let h be the mass matrix of the i-th tire. A,b h is the transformation matrix between the vehicle coordinate system and the geodetic coordinate system. b,u,i This is the transformation matrix between the vehicle coordinate system and the i-th tire coordinate system.

[0044] The kinetic energy of the unsprung mass can be expressed by the following formula:

[0045]

[0046] Among them, T u V is the kinetic energy of the unsprung mass. uz M represents the vertical velocity of the four tires. u Let ω be the mass matrix of the unsprung mass. u J is the rotational angular velocity of the four tires. u Let be the moment of inertia matrix of the unsprung mass.

[0047] The generalized forces acting on the sprung mass and the unsprung mass are respectively expressed by the following formulas:

[0048]

[0049] Where F is the generalized force acting on the vehicle, and r b For, F b Let r be the generalized force matrix acting on the sprung mass. u For, F u This is the matrix of generalized forces acting on unsprung masses.

[0050] The generalized force matrix of the sprung mass can be expressed by the following formula:

[0051]

[0052] Among them, F x,i Let δ be the longitudinal force at the i-th tire contact point with the ground. i For the steering angle input of the i-th tire, F y,i F is the lateral force at the i-th tire contact point with the ground. wx For longitudinal air resistance, F wy For lateral air resistance, F bs,i Let g be the suspension force acting on the i-th tire, g be the acceleration due to gravity, and T be the acceleration due to gravity. d,i M is the torque input for the i-th tire. wx M is the air drag torque about the x-axis. wy Z is the air drag torque about the y-axis. A For the height of the vehicle's center of gravity, M z,i Let x be the return torque of the i-th tire. w,i Let y be the longitudinal position of the i-th tire in the vehicle coordinate system. w,i Let M be the lateral position of the i-th tire in the vehicle coordinate system. wz Let be the air drag torque about the z-axis.

[0053] The generalized force matrix of the unsprung mass can be expressed by the following formula:

[0054]

[0055] Among them, T d,fr T d,fl T d,rr and T d,rl M is the torque input for the right front, left front, right rear, and left rear tires. y,fr M y,fl M y,rr and M y,rl The rolling resistance torques F for the right front, left front, right rear, and left rear tires, respectively. x,fr F x,fl F x,rr and F x,rl The longitudinal forces at the contact points between the right front, left front, right rear, and left rear tires and the ground are respectively, z w,fr z w,fl z w,rr and z w,rl The vertical positions of the centers of gravity of the right front, left front, right rear, and left rear tires are respectively, F z,fr F z,fl F z,rr and F z,rl The vertical forces F acting on the right front, left front, right rear, and left rear tires are respectively. bs,fr F bs,fl F bs,rr and F bs,rl These are the suspension forces acting on the right front, left front, right rear, and left rear tires, respectively, in m. w,fr m w,fl m w,rr and m w,rl The tire masses are those of the front right, front left, rear right, and rear left tires, respectively.

[0056] A2. Based on the dependence of the powertrain mass and output torque on design parameters, a powertrain mass model is constructed. The powertrain mass model includes a motor mass model and a transmission mass model; the motor mass model can be expressed by the following formula:

[0057]

[0058] Where, m d For the mass of the motor, ρ m P is the motor mass factor. max n is the maximum mechanical power of the motor. b This is the reference speed of the motor.

[0059] The mass model of the transmission can be expressed by the following formula:

[0060]

[0061] Where, mgt m is the total mass of all gear sets in the gearbox. g,j Let N be the mass of the j-th pair of gears. g This represents the number of gear sets.

[0062] A3. Based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, simulated control is performed under preset operating conditions to obtain an initial solution. In this embodiment, step A3 specifically involves: based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, constructing a driver model to control the vehicle under the operating conditions of the path tracking model, obtaining a time-series control input and corresponding state output as the initial solution.

[0063] The path tracking model can be represented by the following formula:

[0064]

[0065] Where s is the travel distance of the road, n is the vertical distance of the road, and χ is the centroid sideslip angle. Let X be the road curvature, X be the longitudinal position of the vehicle's center of gravity, θ be the heading angle, θ = ψ - χ, and Y be the lateral position of the vehicle's center of gravity. The road curvature can be expressed by the following formula:

[0066]

[0067] Where dx and ddx are the first and second gradients of the X-coordinate, respectively, and dy and ddy are the first and second gradients of the Y-coordinate, respectively.

[0068] The driver model includes a longitudinal controller and a lateral controller; the longitudinal controller is based on PID control logic, and the lateral controller is based on PD control logic.

[0069] The longitudinal controller of the driver model is shown in the following formula:

[0070]

[0071] Among them, T d e(n) represents the desired driving torque input at the nth time step, where n is the time step size, e(n) is the longitudinal velocity tracking error at the nth time step, and K p T is the proportional gain coefficient of the longitudinal controller. i Let T be the integration time. d The derivative is time.

[0072] The lateral controller of the driver model is shown in the following formula:

[0073] δ(n)=K p1 e1(n)+K d1(e1(n)-e1(n-1))+K p2 e2(n)+K d2 (e2(n)-e2(n-1)).

[0074] Where δ(n) is the desired driving angle input at the nth time step, K p1 and K p2 K is the proportional gain coefficient of the transverse controller. d1 and K d2 Here, e1(n) represents the derivative gain coefficient of the lateral controller, and e2(n) represents the error in vertical distance and the error in yaw angle tracking, respectively.

[0075] A4. Based on the multi-degree-of-freedom vehicle dynamics model and powertrain mass model, construct a multi-objective collaborative optimization problem according to the initial solution, design parameter set, control parameter set and related constraints.

[0076] A5. The multi-objective collaborative optimization problem is transformed into a nonlinear optimization problem, and a nonlinear optimization problem solver is used to solve the problem, obtaining the optimal control parameter set and the optimal design parameter set. Specifically, in this embodiment, the nonlinear optimization problem is as follows:

[0077]

[0078] p = [n b ,β,i g ],n p =3.

[0079] u=[δ,T m1 ,T m2 ,T m3 ,T m4 ],n u =4.

[0080] T w =[T m1 i g ,T m2 i g ,T m3 i g ,T m4 i g ].

[0081] Where J is the objective function value, t f For the terminal time, J p The penalty function is used to constrain acceleration and braking to not be performed simultaneously. w is the weighting factor of the penalty function, f[] is the dynamic model equation, x is the set of state parameters, u is the set of control parameters, t is the time step, p is the set of design parameters, and X... A,b and Y A,bLet be the absolute displacement of the center of mass along the geodetic coordinate system. φ and ψ represent the rotation angles of the vehicle body about the three axes of the geodetic coordinate system, respectively, and z represents the vertical height of the vehicle's center of mass. ufr z ufl z urr and z url θ represents the vertical height of the centers of mass of the four tires, respectively. ufr θ ufl θ urr and θ url Here, represents the rotation angles of the four tires, s represents the distance traveled on the road, n represents the vertical distance traveled on the road, and χ represents the sideslip angle. b β is the reference speed of the motor, β is the high constant power speed ratio of the motor, and i g n is the transmission ratio. p The number of design parameters, δ is the front wheel steering angle of the vehicle, and T is the number of design parameters. m1 T m2 T m3 and T m4 The output torques of the four motors are n, respectively. u To control the number of parameters, T w This represents the torque acting on the four tires.

[0082] In another exemplary embodiment of this application, step A5 may employ the local collocation method, the difference method, and the autoscaling method to transform the multi-objective collaborative optimization problem into a nonlinear optimization problem.

[0083] To better understand the vehicle design optimization method based on collaborative optimal design and optimal control proposed in the above embodiments of this application, the following example uses a four-wheeled independent electric drive vehicle as an example, and the method includes the following steps:

[0084] Step 1: Construction of a multi-degree-of-freedom vehicle dynamics model.

[0085] In this step, based on the vehicle type, a high-precision multi-degree-of-freedom vehicle dynamics model is built using vectorization methods. To more accurately describe the overall dynamic behavior of a four-wheel drive vehicle with independent suspension in cornering, braking, and acceleration, and to facilitate real-time simulation calculations, a multi-degree-of-freedom vehicle dynamics model is built based on Lagrange dynamics. Accurately describing the vehicle's dynamic behavior requires estimating the external forces acting on the vehicle as precisely as possible; therefore, tire forces are calculated using a magic formula tire model. Simultaneously, to improve computational efficiency, vectorized programming is used to establish a multi-degree-of-freedom vehicle dynamics model that supports vectorized calculations.

[0086] like Figure 2As shown, the overall vehicle dynamics model architecture is presented, including the interaction between the vehicle body, suspension, unsprung mass, and tire model. The driver model provides tire steering angle and torque input. In the figure, T... d,i For the torque input of each tire, δ i Input the steering angle for each tire, ω i F represents the rotational speed of each tire. x,i F y,i Vb represents the lateral and longitudinal forces of each tire. i Zb represents the speed at the connection point between the vehicle body and the suspension. i F represents the vertical coordinate of the connection point between the vehicle body and the suspension. sb,i The vertical force exerted by the suspension on the vehicle body is given. The vertical force exerted by the suspension on the tires is equal in magnitude and opposite in direction to the vertical force exerted by the suspension on the vehicle body; therefore, -F is used in the calculation. sb,i D indicates i F represents the vertical displacement of the tire. z,i V is the vertical force exerted on the tire by the ground. u,i Let F be the vector consisting of the lateral and longitudinal velocities of the i-th tire. i ,∑M i This refers to the resultant force and resultant torque acting on the vehicle body in three directions.

[0087] The inputs to the multi-degree-of-freedom vehicle dynamics model are the steering wheel angle and wheel torque. Tire rotation is driven by wheel torque and longitudinal ground force. The tire model inputs are wheel rotational angular velocity and wheel center velocity, and the outputs are tire force and torque. The vertical motion of the unsprung mass is generated by the vertical force of the tire and the suspension force. The motion of the vehicle body is generated by the combined action of longitudinal and lateral tire forces, aerodynamics, and suspension forces.

[0088] In this example, for step A1 in the above method embodiment, the dynamic behavior of a simplified vehicle composed of multiple rigid components is represented by a multi-degree-of-freedom vehicle dynamics model, such as... Figure 3 As shown, the vehicle body can move in three directions: longitudinal, lateral, and vertical, as well as rotate around three axes: x, y, and z. Each of the four tires has two degrees of freedom, namely rotation and vertical motion, which is a 14-DOF vehicle dynamics model.

[0089] In the modeling process, O-XYZ represents the geodetic coordinate system, O b -x b y b z b Representing the vehicle coordinate system, the generalized coordinates with multiple degrees of freedom are represented by vectors as follows:

[0090]

[0091] In the first row q bThe X represents the vehicle's degrees of freedom, namely the absolute displacement of its center of mass along the geodetic coordinate system and its rotation about three axes. A ,Y A Z A This represents the absolute displacement of the centroid along the global coordinate system. φ and ψ represent the rotation angles of the vehicle body around the three axes of the global coordinate system; the second line q u θ represents the rotation and vertical displacement of the four tires. ufr ,θ ufl ,θ urr ,θ url These represent the angular velocity sequences generated by the rotational motions of the four tires: right front, left front, right rear, and left rear. ufr ,z ufl ,z urr ,z url These represent the vertical displacement sequences of the four tires: right front, left front, right rear, and left rear. The velocity vector representation is as follows:

[0092]

[0093] A dot above a variable indicates its derivative. The xy coordinates of the four wheel centers in the vehicle coordinate system are:

[0094]

[0095] Where l represents the longitudinal distance from the vehicle axle to the center of gravity, w represents the wheelbase, and both l and w are distances in the vehicle coordinate system. z represents the relative position of the unsprung mass in the vehicle coordinate system. w It can be represented by its absolute coordinates, roll angle, pitch angle, and xy coordinates in the vehicle coordinate system. The vehicle's xy plane is as follows: Figure 4 As shown, the specific representation is as follows:

[0096]

[0097] The vertical velocity of the unsprung mass in the vehicle coordinate system can be expressed as:

[0098]

[0099] The multi-degree-of-freedom dynamics model of a vehicle can be represented using Lagrange dynamics as follows:

[0100]

[0101] In the formula, T is the kinetic energy of the system, and Q is the kinetic energy of the system. b Q u It refers to the generalized force acting on both sprung and unsprung masses.

[0102] The kinetic energy of the system includes the kinetic energy of the sprung mass and the kinetic energy of the unsprung mass. The solutions for the two are as follows:

[0103] Kinetic energy T of the sprung mass b As shown in the following formula:

[0104]

[0105] Among them, V b V ui These are the components of the generalized velocity of the vehicle body and the velocity of the unsprung mass in the vehicle body coordinate system, respectively. b M ui Let be the mass matrices of the vehicle body and unsprung mass, respectively. There is a coordinate transformation between the vehicle coordinate system and the geodetic coordinate system; the transformation matrix is ​​as follows:

[0106]

[0107] From the above formula, the vehicle speed in the vehicle coordinate system is:

[0108]

[0109] The lateral and longitudinal velocities of the four wheels in the vehicle coordinate system are shown in the following formulas:

[0110]

[0111] The mass matrix of the vehicle body and the mass matrix of the four tires are as follows:

[0112]

[0113] Based on the above calculations, the kinetic energy of the sprung mass can be expressed as:

[0114]

[0115] The kinetic energy of the unsprung mass consists of the rotation and vertical motion of the tire, as shown in the following equation:

[0116]

[0117] Among them, V uz ω u These are the vertical velocity and angular velocity of the four unsprung tires, respectively, M u J u These are the mass matrix and moment of inertia matrix of the unsprung mass, respectively. The calculation method is analogous to that of the sprung mass.

[0118] The generalized forces acting on a vehicle in the dynamic model can be composed of the generalized forces acting on the sprung mass and the unsprung mass. Among them, the vertical forces acting on the sprung mass include gravity, suspension force, and front and rear air pressure; the longitudinal forces include the tire longitudinal force and air resistance; and the lateral forces are only the lateral forces acting on the tires on the ground.

[0119] Vehicle motion depends on the forces and torques applied to the tires; therefore, accurate modeling of tire-road interaction forces is crucial for vehicle dynamics modeling and simulation. Figure 5 The diagram shown below illustrates the tire-road contact. The following section uses the magic formula to accurately model the tire-road interaction forces.

[0120] The effective radius of a wheel is calculated as follows:

[0121]

[0122] Among them, R i,0 Let F be the initial radius of the tire when no external force is applied. z0,i For the nominal tire load, K t,i For tire elastic stiffness, F z,i B is the vertical force exerted on the tire by the ground. reff D reff F reff These represent the effective rolling radii under low load stiffness, peak load stiffness, and high load stiffness, respectively. The vertical force consists of gravity and suspension force.

[0123] F z,i =F bs,i +m u,i g.

[0124] Among them, F z,i F is the vertical force acting on the tire. bs,i For suspension force, m u,i g is the weight of the tire, and the longitudinal slip ratio at the contact point between the tire and the ground is calculated as follows:

[0125]

[0126] The velocity of the wheel center is shown in the following formula:

[0127]

[0128] Tire slip angle calculation:

[0129]

[0130] The longitudinal force at the tire-ground contact point can be described by the magic formula as follows:

[0131] F x,i =(D x,i sin(C x,i arctan(B x,i κ i -E x,i (B x,i κ i -arctan(Bx,i κ i ))))+SV x,i )G xα,i .

[0132] The lateral force at the tire-ground contact point can be described using the magic formula as follows:

[0133] F y,i =(D y,i sin(C y,i arctan(B y,i α i -E y,i (B y,i α i -arctan(B y,i α i ))))+SV y,i )G yκ,i +SV yκ , .

[0134] The tire rollover torque can be described by the magic formula as follows:

[0135]

[0136] The rolling resistance torque of a tire can be described by the magic formula as follows:

[0137]

[0138] The tire self-alignment torque can be described by the magic formula as follows:

[0139] M z =-t(F y -S Vyk )+M zr +sF x .

[0140] Next, we analyze the suspension forces, which consist of spring forces and damping forces. When the stiffness and damping ratio are constant, the suspension forces can be expressed as follows:

[0141]

[0142] The force exerted on a spring is expressed as the spring's stiffness, k. bs,i and deformation Δl s.i The function of the damping force is the damping c. d,i and A function of velocity.

[0143] The spring stiffness can be a constant value or a nonlinear function, while the damping ratio can be a constant or a nonlinear function of the damping velocity, as shown in the following formula. In this case, the spring force and the damping force are nonlinear functions of the deformation and the damping velocity, respectively.

[0144]

[0145] In this example, the nonlinear function is described using tables and one-dimensional interpolation. The deformation Δl of the spring travel... s.i It is the wheel bounce ΔD i The function can be expressed by the transmission ratio λ. s,i calculate:

[0146] Δl s,i =λ s,i ΔD i .

[0147] Among them, the wheel runout ΔD i The vertical displacement of the wheel, or the relative vertical movement of the axle with respect to the vehicle body reference frame, can be defined as:

[0148] ΔD i =z w,i -z w0,i .

[0149] The deformation rate of the suspension damper can be denoted as:

[0150]

[0151] Among them, z w,i Let z be the vertical position of the unsprung mass in the vehicle coordinate system. w0,i Let these be the initial values. Finally, the suspension force acting on each wheel can be expressed as:

[0152] F bs,i =λ s,i F bss,i .

[0153] Next, we derive the air resistance model. First, we calculate the air resistance sideslip angle. Assuming the air is still and the wind speed is 0, the air resistance sideslip angle is the same as the vehicle's center of gravity sideslip angle, denoted as:

[0154]

[0155] Then, the air drag coefficient is calculated based on the air drag sideslip angle using linear interpolation and cubic spline interpolation.

[0156]

[0157] Calculate the air drag factor:

[0158]

[0159] Substituting the above results, the air resistance can be calculated as follows:

[0160]

[0161] Vehicle force diagram as follows Figure 6 As shown. Therefore, the generalized force matrix of the sprung mass is as follows:

[0162]

[0163] Forces and moments acting on unsprung mass as follows Figure 7 As shown, its generalized force matrix is ​​expressed as:

[0164]

[0165] The forces and moments in the formula have been solved in the previous text, δ i Input for tire steering angle, T d,i Based on force analysis and the principle of virtual work, the generalized force can be derived from the torque input of each tire, as shown in the following equation:

[0166]

[0167] Based on Lagrange mechanics and d'Alembert's principle, the generalized equations of motion for rigid and unsprung masses are derived in the form of the following equations. Directly using the differential generalized mass matrix can improve computational efficiency.

[0168]

[0169] Step 2: Construction of the powertrain mass model.

[0170] In this example, regarding step A2 in the above method embodiment, considering that transmission system quality modeling mainly focuses on the dependence of transmission system quality on design parameters, a quality model of the motor and gearbox is built to facilitate vehicle lightweight design.

[0171] First, the motor mass model is derived: In this example, an alternating current (AC) motor is chosen because of its high output power and light weight. Based on the fundamental theory of AC motor design, the relationship between mass and maximum power P is derived. max The relationship between the basic speed n and the rotational speed n. The following formula is the basic reference equation for designing AC motors:

[0172]

[0173] Where P' is the apparent power, N pHere, E is the number of phases, I is the induced electromotive force, f is the current frequency, N is the number of series turns of the stator, and K is the current frequency. w It is the winding coefficient, Φ m B is the maximum magnetic flux per pole, and A is the magnetic flux density through the iron. Fe J is the cross-sectional area of ​​the iron, and A is the current density. Cu It is the total cross-sectional area of ​​the winding coil, p is the number of pole pairs, N b This is the basic speed of the motor. It is determined by the power factor cosφ and efficiency η. m From the apparent power P', we can obtain the maximum mechanical power P of the motor. max :

[0174]

[0175] In the formula, K E Cosφ is the ratio of electromotive force to terminal voltage, and cosφ is the power factor, which is typically 0.85 for induction motors.

[0176] Based on the above equations, the relationship between the power and mass of the motor can be derived. The cross-sectional area A of the iron... Fe The total cross-sectional area A of the winding coil Cu Each is proportional to the square of the basic unit "length":

[0177]

[0178] Volume is proportional to the cube of the length of the basic unit:

[0179] V d ∝l 3 .

[0180] Based on the above equation, the mass of the motor can be expressed as:

[0181]

[0182] Where, ρ m This is the motor mass factor.

[0183] Next, we derive the transmission mass model: the transmission is one of the heaviest components in the powertrain; therefore, reducing its weight is a primary consideration in vehicle lightweight design. We apply the fundamental equations relating gear dimensions, transmission power, input shaft gear speed, and gear ratio of helical / spur gear sets:

[0184]

[0185] Where, d sis the center distance between the two shafts (mm), w is the gear width (mm), assuming all gears have the same width, P is the transmission power (kW), n is the input gear speed (rpm), K is the surface durability coefficient (N / mm2), and i is the gear ratio.

[0186] When the physical units of power and rotational speed are kW and Rpm respectively, the torque on the input shaft Tin can be expressed as:

[0187]

[0188] Therefore, we can conclude that:

[0189]

[0190] The gear ratio and center distance can be expressed as:

[0191]

[0192] From the above two equations, we can deduce that:

[0193]

[0194] d in d is the diameter of the input gear. o Let the diameter of the output gear be denoted as , then substituting it into the above formula, we get:

[0195]

[0196] The mass of a gear set can then be expressed as:

[0197]

[0198] ψ is the gear volume fill factor, ρ is the gear mass density, and combining the above equations, we can obtain:

[0199]

[0200] Therefore, the total mass of all gear sets in the entire gearbox can be estimated as:

[0201]

[0202] Where, m g For the mass of a pair of gear sets, m gt This represents the total mass of all gear sets.

[0203] Step 3: Model and simulate the working conditions to obtain the initial solution.

[0204] In this example, for step A3 in the above method embodiment, the preset working condition is taken as the Nürburgring circuit. The XY coordinates of the example circuit are as follows: Figure 8 As shown. This data can be obtained via GPS or extracted from commercial or open-source maps.

[0205] The curvature of a road can be calculated using given XY coordinates:

[0206]

[0207] In a curvilinear coordinate system, a road can be described by its curvature and arc length, such as... Figure 9 As shown, this represents the trajectory curvature of the example track, with the origin at [0,0]. Here, dx, ddx, dy, and ddy are the first and second order gradients of the X and Y coordinates, respectively.

[0208] In a curvilinear coordinate system, the vehicle's position on the road can be described by the distance traveled, *s*, the distance *n* from the reference trajectory (normal), and the heading angle *θ* at the current distance traveled. Figure 10 As shown, the heading angle is denoted as:

[0209] θ = ψ - χ.

[0210] The derivative of the travel distance s can be expressed by the vehicle's longitudinal and lateral velocities in the global coordinate system:

[0211]

[0212] The derivative of the vertical distance n can be expressed as:

[0213]

[0214] The derivative of the centroid sideslip angle χ can be expressed as:

[0215]

[0216] In addition, a driver model is developed to control the vehicle to travel along a reference trajectory on a given road. The driver model is divided into longitudinal control and lateral control.

[0217] The purpose of longitudinal control is to make the vehicle track a given reference longitudinal speed, which is based on PID control logic:

[0218]

[0219] In the formula, T d (n) represents the desired driving torque, n is the time step, e is the longitudinal speed tracking error, and K p For proportional gain, T i Let T be the integration time. d The derivative is time.

[0220] The purpose of lateral control is to generate an appropriate steering angle to control the vehicle to follow a reference trajectory. Considering the vehicle's delayed response, a lateral controller based on PD control is proposed to accurately track a given reference trajectory.

[0221] δ(n)=K p1 e1(n)+K d1 (e1(n)-e1(n-1))+K p2 e2(n)+K d2 (e2(n)-e2(n-1)).

[0222] In the formula, K p1 K p2 For proportional gain, K d1 K d2 Here, e1 and e2 represent the vertical distance and yaw angle tracking errors, respectively.

[0223] e1 = n, e2 = ψ ref -ψ.

[0224] Step 4: Constructing the multi-objective collaborative optimization problem.

[0225] In this example, for step A4 in the above method embodiment, the following method is used: Figure 11 The cooperative optimal design and optimal control algorithm framework shown above, based on the vehicle dynamics model established earlier in this example, provides the input to the problem. The input to the problem includes at least the initial guess value x0 of the state variables, the initial guess value u0 of the control variables, and the terminal time t. f0 And design variable p0.

[0226]

[0227] Where, N x,usr It is the number of nodes for the user-provided state variables, where t is the number of nodes from 0 to t. f Discretized into N x,usr Time series of N points, u,usr It is the number of nodes for the user-provided control variables, n. x The number of state variables n in the dynamic equation u This refers to the number of control variables. If the terminal time is also a parameter to be optimized, then n... tf It is 1, otherwise it is 0, n p It refers to the number of design parameters. Let represent a two-dimensional matrix with i rows and j columns. For the global transformation method, the initial values ​​of the start and end times for each stage will be generated by the optimal control framework.

[0228] The upper and lower bounds of the state variables, control variables, terminal time, and design variables, as well as the inequality constraint g, are given by the following formula, ng Number of path constraints:

[0229]

[0230] In addition to the initial values ​​and upper and lower bound constraints mentioned above, the problem input also includes the Lagrange term of the cost function. Meyer The first-order dynamic constraint function f, the path constraint function g, and the boundary constraint function b, with their input and output dimensions shown in the following equation:

[0231]

[0232] In addition to the parameters mentioned above, the input for the problem also includes the parameters listed in the table below:

[0233] Table 1 Other input parameters for the optimization problem

[0234]

[0235]

[0236] Step 5: Transform the problem and solve it.

[0237] In this example, regarding step A5 of the above method embodiment, in the problem transformation section, the format of the variables and functions in the problem input is transformed into the format required by the NLP solver. This section can use methods such as local collocation, finite difference, and automatic scaling. The problem transformation includes the following steps:

[0238] Step 5.1: In the local collocation method, the continuous state and control variables over the entire time interval can be discretized into N using linear interpolation. n The new variables obtained from the given nodes are given by the following formula:

[0239]

[0240] Then, the discrete state variables, control variables, and static variables are reconstructed into NLP column vectors:

[0241]

[0242] For the Hermite-Simpson method, when the control variables in each sampling interval are also selected as parameters to be optimized, the reconstructed NLP variables are:

[0243]

[0244] When the state variables in each sampling interval are also selected as parameters to be optimized, the reconstructed NLP variables are as follows:

[0245]

[0246] The upper and lower bounds of the variables mentioned above should also be reconstructed based on the reconstructed NLP variables.

[0247] Step 5.2: Constraints.

[0248] 1) Residual constraints of the Hermite-Simpson method.

[0249] Based on third-order Hermite interpolation, in x k With x k+1 State variables between and its derivative The following can be derived:

[0250]

[0251] Based on the Hermite-Simpson method, in residual constraints at the location It can be represented as:

[0252]

[0253] Apply linear interpolation to calculate x k With x k+1 Control variables between

[0254]

[0255] To reduce redundant calculations in numerical computation, matrix calculations are applied; therefore, state variables... and control variables It can be represented as:

[0256]

[0257] The system's dynamic equations are:

[0258]

[0259] Transformation matrix T s1 and T s2 for:

[0260]

[0261] The residual constraint ζ can be represented by matrix calculation as follows:

[0262]

[0263] 2) Residual constraints of the Trapezoidal method.

[0264] Residual constraint ζ of the Trapezoidal method k It can be represented as:

[0265]

[0266] The matrix calculation is as follows:

[0267]

[0268] 3) Path constraints and boundary constraints.

[0269] Path constraints are functions of state variables, control variables, design variables, and time; boundary constraints are functions of the initial state and the final state.

[0270]

[0271] Finally, after all constraint calculations are completed, the NLP constraints are represented as:

[0272]

[0273] All residual constraints have upper and lower bounds of 0. The upper and lower bounds of path constraints and boundary constraints have been given previously.

[0274] Step 5.3: Jacobian matrix.

[0275] 1) Jacobian matrix of residual constraints

[0276] Residual constraints when the state and control variables at the midpoint of each discrete time interval are not considered It is x a ,u a ,x b ,u b ,t f A function of p, x a ,u a ,x b ,u b ,t f The specific values ​​of p are as follows:

[0277]

[0278] To avoid redundant calculations, matrix calculations are performed. First, the derivative of ζ with respect to x, u, and p is calculated:

[0279]

[0280] Therefore, the Jacobian matrix can be simplified to:

[0281]

[0282] In the formula Let T be a matrix where the i-th column is 1 and all other elements are 0. a ,T b It is given by the following formula:

[0283] 2) Jacobian matrix of path constraints and boundary constraints

[0284] First, calculate the derivative of the path constraint g with respect to x, u, p, t:

[0285]

[0286] Then calculate the boundary constraint b for x0, x f The derivatives of p and t:

[0287]

[0288] Step 5.4: Solve for the cost function.

[0289] The input cost function consists of Meyer and Lagrange terms, and can be expressed as:

[0290]

[0291] Step 5.5: Solve for the gradient of the cost function.

[0292] The cost function is [x,u,t] f The function is defined as [p], therefore its gradient with respect to the state variable and the control variable is as follows:

[0293] Then, the gradient of the cost function with respect to the terminal time and design variables is calculated, and finally the NLP gradient is obtained:

[0294]

[0295] After the above problem transformation is performed, the transformed problem is input into the underlying solver for solving, which can simultaneously solve for the optimal control parameters and the optimal design parameters.

[0296] The nonlinear optimization problem takes the form shown below:

[0297]

[0298] State variables: Design variables: p = [n] b ,β,i g ],n p =3, Control variables: u=[δ,Tm1 ,T m2 ,T m3 ,T m4 ],n u =4, Tire torque: T w =[T m1 i g ,T m2 i g ,T m3 i g ,T m4 i g ].

[0299] After the above input is provided, the problem is transformed. An NLP problem solver is then used to solve the transformed problem, yielding the time-series vehicle state variables and control variables, as well as the optimized shortest time and corresponding optimal design variables. In practical applications, this is divided into two parts: the optimized design variables can be used for vehicle development, such as motor parameter selection; the obtained state variables and control variables can be used for developing control algorithms.

[0300] In the above method embodiments, this application constructs a vehicle-oriented cooperative optimal design and optimal control framework, which can simultaneously optimize the design parameters and control parameters based on the vehicle dynamic model. The optimal design parameters can be used for vehicle design, and the optimal control parameters can serve as a reference for subsequent online control strategy formulation. This framework can apply gradient information to expand the optimization space and has good optimization performance. This is mainly related to the construction of the optimal design and optimal control problem, the rationality of vehicle dynamics modeling, and the efficient solution of the problem. In this embodiment, for large-scale multi-degree-of-freedom optimal design and optimal control problems, the stability and efficiency of the solution are greatly improved. This is mainly due to the vehicle vectorized dynamics modeling method, and in the solution part of the optimal problem, the direct collocation method is improved, and an automatic normalization algorithm is applied, realizing a stable and fast solution to large-scale cooperative optimal problems.

[0301] Based on the same inventive concept, this application also provides an apparatus for implementing the vehicle design optimization method of cooperative optimal design and optimal control as described above. The solution provided by this apparatus is similar to the implementation scheme described in the above method; therefore, the specific limitations in one or more apparatus embodiments provided below can be found in the limitations of the vehicle design optimization method of cooperative optimal design and optimal control described above, and will not be repeated here.

[0302] In one exemplary embodiment, such as Figure 12 As shown, a vehicle design optimization device for collaborative optimal design and optimal control is provided, comprising the following modules:

[0303] The dynamics model building module is used to construct multi-degree-of-freedom vehicle dynamics models based on Lagrange dynamics, according to the vehicle type.

[0304] The mass model building module is used to construct a powertrain mass model based on the dependence of the powertrain mass and output torque on design parameters.

[0305] The initial solution generation module is used to perform simulation control under preset working conditions based on a multi-degree-of-freedom vehicle dynamics model and a powertrain mass model to obtain the initial solution.

[0306] The multi-objective optimization problem construction module is used to construct multi-objective collaborative optimization problems based on multi-degree-of-freedom vehicle dynamics models and powertrain mass models, according to the initial solution, design parameter set, control parameter set, and relevant constraints.

[0307] The problem transformation and NLP solving module is used to transform multi-objective collaborative optimization problems into nonlinear optimization problems, and to solve the problems using a nonlinear optimization problem solver to obtain the optimal control parameter set and the optimal design parameter set.

[0308] certainly, Figure 12 The architecture shown is merely exemplary; it can be omitted as needed when implementing different functionalities. Figure 12 One or at least two components of the system shown.

[0309] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0310] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A vehicle design optimization method based on collaborative optimal design and optimal control, characterized in that, include: Based on the vehicle type, a multi-degree-of-freedom vehicle dynamics model is constructed using Lagrange dynamics. Based on the dependence of the powertrain mass and output torque on design parameters, a powertrain mass model is constructed. Based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, simulation control is performed under preset working conditions to obtain an initial solution; Based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, a multi-objective collaborative optimization problem is constructed according to the initial solution, design parameter set, control parameter set and relevant constraints. The multi-objective collaborative optimization problem is transformed into a nonlinear optimization problem, and the problem is solved using a nonlinear optimization problem solver to obtain the optimal control parameter set and the optimal design parameter set. The multi-degree-of-freedom vehicle dynamics model can be expressed by the following equation: ； ； in, For the kinetic energy of the vehicle's power system, The generalized force on the sprung mass For the generalized force on unsprung mass, Indicates the degrees of freedom of the vehicle body. This represents the rotation and vertical displacement of the four tires. X A , Y A and Z A These represent the absolute displacements of the centroid along the geodetic coordinate system. These represent the rotation angles of the vehicle body around the three axes of the geodetic coordinate system. These represent the angular velocity sequences generated by the rotational motion of the four tires: right front, left front, right rear, and left rear. These represent the vertical displacement sequences of the four tires: right front, left front, right rear, and left rear. The powertrain mass model includes a motor mass model and a transmission mass model; the motor mass model can be expressed by the following formula: ； in, m d For motor quality, ρ m For motor mass factor, P max This is the maximum mechanical power of the motor. n b This is the reference speed of the motor; The mass model of the transmission can be expressed by the following formula: ； in, m gt The total mass of all gear sets in the gearbox. m g,j For the first j Regarding the mass of the gear set, N g This represents the number of gear sets.

2. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 1, characterized in that, The kinetic energy of a vehicle's powertrain includes the kinetic energy of the sprung mass and the kinetic energy of the unsprung mass; the kinetic energy of the sprung mass can be expressed by the following formula: ； in, T b The kinetic energy of the sprung mass. V b Let be the component of the generalized velocity of the vehicle body in the vehicle body coordinate system. M b Here is the mass matrix of the vehicle body. V u,i For the first i The speed of each tire in the vehicle's coordinate system M u,i For the first i The mass matrix of each tire h A,b This is the transformation matrix between the vehicle coordinate system and the geodetic coordinate system. h b,u,i For the vehicle body coordinate system and the first i Transformation matrix between tire coordinate systems; The kinetic energy of the unsprung mass can be expressed by the following formula: ； in, T u The kinetic energy of the unsprung mass. V uz The vertical velocities of the four tires are... M u The mass matrix is ​​the unsprung mass. ω u The rotational angular velocity of the four tires. J u Let be the moment of inertia matrix of the unsprung mass.

3. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 1, characterized in that, The generalized forces acting on the sprung mass and the unsprung mass are respectively expressed by the following formulas: ； in, F The generalized forces acting on a vehicle r b for, F b The generalized force matrix is ​​the matrix of forces acting on the sprung mass. r u for, F u This is the matrix of generalized forces acting on unsprung masses; The generalized force matrix of the sprung mass can be expressed by the following formula: ； in, F x,i For the first i The longitudinal force at the contact point between the tire and the ground For the first i Input the steering angle of each tire. F y,i For the first i Lateral force at each tire contact point with the ground F wx For longitudinal air resistance, F wy For lateral air resistance, F bs,i To act on the first i Suspension force on each tire g It is the acceleration due to gravity. T d,i For the first i Torque input to each tire, M wx To bypass x Air drag torque of the shaft, M wy To bypass y Air drag torque of the shaft, Z A The height of the vehicle's center of gravity, M z,i For the first i The self-aligning torque of each tire x w,i For the first i The longitudinal position of each tire in the vehicle coordinate system y w,i For the first i The lateral position of each tire within the vehicle's coordinate system M zy To bypass z Air drag torque of the shaft; The generalized force matrix of the unsprung mass can be expressed by the following formula: ； in, T d,fr , T d,fl , T d,rr and T d,rl Torque input for the right front, left front, right rear, and left rear tires. M y,fr , M y,fl , M y,rr and M y,rl These are the rolling resistance torques of the right front, left front, right rear, and left rear tires, respectively. F x,fr , F x,fl , F x,rr and F x,rl The longitudinal forces at the contact points between the right front, left front, right rear, and left rear tires and the ground. z w,fr , z w,fl , z w,rr and z w,rl These are the vertical positions of the centers of gravity of the right front, left front, right rear, and left rear tires, respectively. F z,fr , F z,fl , F z,rr and F z,rl The vertical forces acting on the right front, left front, right rear, and left rear tires. F bs,fr , F bs,fl , F bs,rr and F bs,rl The suspension forces acting on the right front, left front, right rear, and left rear tires. m w,fr , m w,fl , m w,rr and m w,rl The tire masses are those of the front right, front left, rear right, and rear left tires, respectively.

4. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 1, characterized in that, Based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, simulation control is performed under preset working conditions to obtain an initial solution. Specifically, based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, a driver model is constructed to control the vehicle to drive under the working conditions of the path tracking model, resulting in a time-series control input and corresponding state output, which serve as the initial solution.

5. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 4, characterized in that, The path tracking model can be represented by the following formula: ； in, s The distance traveled along the road. n The vertical distance of the road. The sideslip angle is the angle of the centroid. For road curvature, X The longitudinal position of the vehicle's center of gravity. θ For heading angle, , Y The lateral position of the vehicle's center of gravity; the road curvature can be expressed by the following formula: ； in, dx and ddx These are the first and second gradients of the X-coordinate, respectively. dy and ddy These are the first and second gradients of the Y-coordinate, respectively; The driver model includes a longitudinal controller and a lateral controller; the longitudinal controller is based on PID control logic, and the lateral controller is based on PD control logic. The longitudinal controller of the driver model is shown in the following formula: ； in, T d ( n ) is the first n The desired driving torque input at each time step n For time step, e ( n ) is the first n Tracking error of longitudinal velocity at each time step K p The proportional gain coefficient of the longitudinal controller. T i For integration time, T d The derivative time; The lateral controller of the driver model is shown in the following formula: ； in, For the first n The expected driving angle input at each time step K p1 and K p2 This is the proportional gain coefficient of the transverse controller. K d1 and K d2 The derivative gain coefficient of the transverse controller. e 1( n )and e 2( n These represent the errors in vertical distance and yaw angle tracking, respectively.

6. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 1, characterized in that, The nonlinear optimization problem is shown in the following equation: ； ； ； ； ； in, J The objective function value, t f For terminal time, J p This is a penalty function used to constrain acceleration and braking to prevent them from occurring simultaneously. w Let f[] be the weighting factor of the penalty function, f[] be the dynamic model equation, x be the set of state parameters, and u be the set of control parameters. t Let p be the time step, and p be the set of design parameters. X A,b and Y A,b Let be the absolute displacement of the center of mass along the geodetic coordinate system. These represent the rotation angles of the vehicle body around the three axes of the geodetic coordinate system. z The vertical height of the vehicle's center of gravity. z ufr , z ufl , z urr and z url The vertical height of the centers of mass of the four tires. θ ufr , θ ufl , θ urr and θ url The rotation angles of the four tires. s The distance traveled along the road. n The vertical distance of the road. The sideslip angle is the angle of the centroid. n b This is the reference speed of the motor. β For high constant power speed ratio of motor. i g The transmission ratio is... n p The number of design parameters, For the steering angle of the vehicle's front wheels, T m1 , T m2 , T m3 and T m4 The output torque of the four motors, n u To control the number of parameters, T w This represents the torque acting on the four tires.

7. The vehicle design optimization method based on collaborative optimal design and optimal control according to claim 1, characterized in that, The multi-objective collaborative optimization problem is transformed into a nonlinear optimization problem by employing the local collocation method, the difference method, and the autoscaling method.

8. A vehicle design optimization device for collaborative optimal design and optimal control, characterized in that, The vehicle design optimization method for implementing the cooperative optimal design and optimal control as described in any one of claims 1-7, wherein the vehicle design optimization apparatus for cooperative optimal design and optimal control comprises: The dynamics model building module is used to construct a multi-degree-of-freedom vehicle dynamics model based on Lagrange dynamics according to the vehicle type. The mass model building module is used to construct a powertrain mass model based on the dependence of the powertrain mass and output torque on design parameters. The initial solution generation module is used to perform simulation control under preset working conditions based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model to obtain an initial solution. A multi-objective optimization problem construction module is used to construct a multi-objective collaborative optimization problem based on the multi-degree-of-freedom vehicle dynamics model and the powertrain mass model, according to the initial solution, design parameter set, control parameter set and related constraints. The problem transformation and NLP solving module is used to transform the multi-objective collaborative optimization problem into a nonlinear optimization problem, and to solve the problem using a nonlinear optimization problem solver to obtain the optimal control parameter set and the optimal design parameter set.

Images