Intelligent networked automobile chassis drive-by-wire simulation method and system based on multi-domain modeling

Abstract

The invention provides an intelligent networked automobile chassis drive-by-wire simulation method and system based on multi-domain modeling, and relates to the technical field of adaptive control, and the method comprises the steps: obtaining structural parameters and kinetic parameters of an intelligent networked automobile chassis, constructing a finite element grid, building a multi-flexible-body kinetic model corresponding to the intelligent networked automobile chassis, and constructing a multi-flexible-body dynamic model corresponding to the intelligent networked automobile chassis; introducing a coupling coordination equation, establishing a multi-flexible-body kinetic equation, and solving to obtain vibration response and stress-strain distribution under different working conditions; generating a plurality of finite element models, performing multi-physical field fusion and multi-error source fusion models, performing online error identification and active compensation, performing all-condition drive-by-wire chassis closed-loop control, and generating a drive-by-wire chassis closed-loop integrated simulation model; constructing a road three-dimensional parameterized model, generating a vehicle-road coupling model, generating a simulation test case, identifying a performance short board, constructing a knowledge graph and an optimization rule base, and generating an optimal control strategy.

Description

Technical Field

[0001] The present invention relates to the field of adaptive control technology, and in particular to a method and system for simulating chassis control-by-wire of an intelligent networked vehicle based on multi-domain modeling. Background Art

[0002] As the automotive electronic and electrical architecture develops towards a centralized structure, the chassis-by-wire technology is increasingly becoming one of the key technologies for intelligent connected vehicles. The chassis-by-wire replaces the traditional mechanical connection and transmits the driver's control instructions to the actuator through electronic signals to achieve control of chassis functions such as steering, braking, and suspension. Compared with the traditional chassis, the chassis-by-wire has higher flexibility, scalability, and intelligence.

[0003] However, the working environment of the drive-by-wire chassis is very complex and needs to adapt to a variety of road types and environmental conditions. Traditional drive-by-wire chassis development mainly relies on experiments and experience. Road tests and bench tests are used to verify and optimize products, which often requires the production of multiple prototypes and a large amount of manpower and material resources. The development cycle is long and the cost is high, making it difficult to fully evaluate the adaptability of the drive-by-wire chassis under complex working conditions.

[0004] Therefore, a solution is urgently needed to solve the problems existing in the prior art. Summary of the invention

[0005] The embodiments of the present invention provide a method and system for intelligent connected vehicle chassis control-by-wire simulation based on multi-domain modeling, which can at least solve some of the problems existing in the prior art.

[0006] A first aspect of an embodiment of the present invention provides a method for simulating chassis control-by-wire of an intelligent connected vehicle based on multi-domain modeling, comprising:

[0007] Acquire structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, introduce a coupling coordination equation at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model through the Lagrange multiplier method and establish a multi-flexible body dynamics equation, solve the multi-flexible body dynamics equation through the Newmark implicit explicit algorithm, and obtain vibration response and stress-strain distribution under different working conditions;

[0008] The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, and multiple finite element models are generated. Multi-physics field fusion is performed at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, and a multi-physics field model is generated and integrated into the drive-by-wire chassis integrated simulation model. A multi-error source fusion model corresponding to multiple sensors is constructed, and online error identification and active compensation are performed through the Kalman filter algorithm to generate sensor data fusion results. The sensor data fusion results, vibration responses and stress-strain distributions under different working conditions are fed back to the controller, and closed-loop control of the drive-by-wire chassis under all working conditions is performed in combination with the dynamic response characteristics of the drive-by-wire chassis to generate a closed-loop integrated simulation model of the drive-by-wire chassis.

[0009] Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parameterized model of the road, embed the three-dimensional parameterized model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model, and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full working condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

[0010] In an optional embodiment,

[0011] The structural parameters and dynamic parameters of the intelligent networked vehicle chassis are obtained, a finite element mesh is constructed based on a parametric finite element method of material nonlinearity and collective nonlinearity, the finite element mesh is adaptively optimized, and a multi-flexible body dynamic model corresponding to the intelligent networked vehicle chassis is established, wherein the multi-flexible body dynamic model includes a flexible body dynamic sub-model and a rigid body dynamic sub-model, and at the coupling interface between the flexible body dynamic sub-model and the rigid body dynamic sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamic equation is established, and the multi-flexible body dynamic equation is solved by a Newmark implicit explicit algorithm, and the vibration response and stress-strain distribution under different working conditions are obtained, including:

[0012] Reconstruct the surface profile information of different components in the intelligent networked vehicle chassis through three-dimensional scanning and structured light through reverse engineering methods to obtain a three-dimensional mesh model, and obtain the structural parameters corresponding to the intelligent networked vehicle chassis based on the three-dimensional mesh model, wherein the structural parameters include the size parameters, posture parameters and connection parameters corresponding to different components, and obtain the dynamic parameters corresponding to the intelligent networked vehicle chassis, wherein the dynamic parameters include mass, stiffness, damping and moment of inertia;

[0013] Based on the structural parameters and the dynamic parameters, a finite element mesh corresponding to the chassis of the intelligent networked vehicle is constructed based on a parametric finite element method that takes into account material nonlinearity and collective nonlinearity; based on topological parameters corresponding to chassis configuration features, a logical relationship between chassis components and interfaces is determined and a consistent topological description is formed; based on the consistent topological description and the finite element mesh, corresponding gradient information of stress and strain is obtained through post-processing analysis; the gradient information is used as an error estimation indicator; the area and degree of local mesh refinement and local mesh coarsening are adaptively determined; a mesh refinement criterion and a mesh coarsening criterion are generated at the same time; and adaptive mesh optimization is performed based on the mesh refinement criterion and the mesh coarsening criterion;

[0014] A multi-flexible body dynamics model corresponding to the intelligent connected vehicle chassis is established based on the optimized finite element mesh, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics model, the flexible body dynamics sub-model determines the large deformation and large rotation corresponding to the flexible body according to the absolute node coordinate method, constructs an initial flexible body dynamics equation in combination with non-conservative forces and corresponding constraints, and performs dimensionality reduction through the modal superposition method to obtain the flexible body dynamics equation, the rigid body dynamics model is modeled through the Newton-Euler equations, defines the position vector of any point of the rigid body and determines the kinetic energy of the rigid body, establishes the translation equation and the rotation equation of the rigid body in combination with the Euler angle parameterized rotation matrix, and determines the rigid body dynamics equation based on the translation equation and the rotation equation;

[0015] Based on the rigid body dynamics equation and the flexible body dynamics equation, the coupled displacement coordination conditions and velocity coordination conditions are defined by the Lagrange multiplier method, the augmented Lagrange equation is constructed in combination with the Lagrange multiplier, the partial derivatives of the generalized coordinates are calculated and the multi-flexible body dynamics equation is constructed, the multi-flexible body dynamics equation is solved by the Newmark implicit and explicit algorithm combined with the mixed integration strategy, and the vibration response and stress-strain distribution of the connected vehicle chassis under different working conditions are obtained.

[0016] In an optional embodiment,

[0017] The multi-flexible body dynamics equation is solved by using the Newmark implicit and explicit algorithm combined with the mixed integration strategy as shown in the following formula:

[0018]

[0019] Among them, q n+1 represents the generalized coordinate vector of the next time step n+1, q n Represents the generalized coordinate vector of the current time step, △t n Represents the adaptive step size of the current time step, q′ n Represents the velocity vector of the current time step, q″ n represents the acceleration vector of the current time step, q″(n+1) represents the acceleration vector of the next time step, and β represents the acceleration control parameter.

[0020] In an optional embodiment,

[0021] The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, and multiple finite element models are generated. Multi-physics field fusion is performed at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, and a multi-physics field model is generated and integrated into the drive-by-wire chassis integrated simulation model. A multi-error source fusion model corresponding to multiple sensors is constructed, and online error identification and active compensation are performed through the Kalman filter algorithm to generate sensor data fusion results. The sensor data fusion results, vibration responses and stress-strain distributions under different working conditions are fed back to the controller, and closed-loop control of the drive-by-wire chassis under all working conditions is performed in combination with the dynamic response characteristics of the drive-by-wire chassis. The generation of a closed-loop integrated simulation model of the drive-by-wire chassis includes:

[0022] The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, an electromagnetic field finite element model is constructed by combining the Maxwell equations and the geometric structure of the motor, the electromagnetic field equations are solved by a meshing method to obtain the magnetic field distribution data inside the motor, a fluid field finite element model is constructed according to the fluid flow and fluid control equations inside the hydraulic cylinder, and the fluid field equations are solved to obtain the velocity field and pressure field corresponding to the fluid, solid mechanics control equations are selected according to the solid structure part of the actuator and a solid field finite element model is constructed, and the solid field equations are solved in combination with the geometric shape and load conditions of the solid structure part to obtain the displacement field and stress field of the solid structure part;

[0023] Based on the electromagnetic field finite element model, the fluid field finite element model and the solid field finite element model, a coupling interface between an actuator and a transmission mechanism is determined, coupling conditions are defined at the coupling interface, a mathematical constraint relationship of the coupling interface is established by a penalty function method, each finite element model is solved at each time step based on the mathematical constraint relationship, multi-physics field fusion is achieved by information exchange and iterative correction at the coupling interface, and the multi-physics field model is obtained and integrated into the integrated simulation model of the drive-by-wire chassis;

[0024] A corresponding measurement model is constructed based on the preset sensors, the sensor error sources are determined based on the measurement model, and the multi-error source fusion model is constructed through the Kalman filter algorithm, the measurement data of each sensor is acquired in real time, and online error identification and active compensation are performed according to the multi-error source fusion model, the data after active compensation are used as the sensor data fusion result, combined with the vibration response and stress-strain distribution feedback and the dynamic response characteristics of the drive-by-wire chassis under the different working conditions, the closed-loop control of the drive-by-wire chassis under all working conditions is performed through a preset controller, and the closed-loop integrated simulation model of the drive-by-wire chassis is obtained.

[0025] In an optional embodiment,

[0026] The mathematical constraint relationship of the coupling interface is established by the penalty function method, and each finite element model is solved at each time step based on the mathematical constraint relationship as shown in the following formula:

[0027]

[0028] Among them, ε total is the overall energy function value, which represents the sum of the finite element model energy and penalty function, ε A is the energy of finite element model A, ε B represents the energy of the finite element model B, k i represents the penalty factor at level i, u A represents the displacement vector of the finite element model A on the coupling interface, u B represents the displacement vector of the finite element model B on the coupling interface, is an indicator function, indicating that when the displacement difference on the coupling interface is greater than the threshold ζ i 1 when ζ is set, otherwise 0. i Represents the displacement difference threshold of the i-th level.

[0029] In an optional embodiment,

[0030] Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parametric model of the road, embed the three-dimensional parametric model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full-condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis, including:

[0031] Collecting multi-source heterogeneous road map data corresponding to different cities, wherein the multi-source heterogeneous road map data includes high-precision maps, laser point clouds and street view images, obtaining the spatiotemporal references of different data sources based on inertial navigation technology and performing point cloud registration through control points, generating consistent point cloud data, and preprocessing the consistent point cloud data through a voxel filtering algorithm to obtain preprocessed multi-source heterogeneous road map data;

[0032] For the pre-processed multi-source heterogeneous road map data, an end-to-end segmentation model is selected to construct a labeled data set for road semantic segmentation, wherein the labeled data set includes road images under different weather and lighting conditions, pixel-level annotation truth values ​​are generated by combining manual annotation and automatic generation, training is performed in combination with a cross entropy loss function and the end-to-end segmentation model is optimized by an optimizer, and a performance evaluation is performed using a pre-set independent test set. If the end-to-end segmentation model meets the pre-set performance standard, the end-to-end segmentation model is used to perform semantic segmentation on the processed multi-source heterogeneous road map data, identify urban roads, and extract road center lines and boundary lines as geometric features through a feature extraction algorithm, estimate road surface roughness and road surface material based on the intensity information of point cloud reflection and use them as road surface attribute parameters, extract environmental information corresponding to different roads based on the street view images, and extract environmental parameters through a feature extraction algorithm, and construct a three-dimensional parameterized road model based on the geometric features, road surface attribute parameters, and environmental parameters, and embed the three-dimensional parameterized road model into the closed-loop integrated simulation model of the drive-by-wire chassis to generate a vehicle-road coupling model;

[0033] Based on the vehicle-road coupling model, the high-dimensional parameter space corresponding to the three-dimensional parameterized model of the road is explored through an intelligent sampling method, and the influence of different dimensional parameters on vehicle performance is evaluated in combination with a sensitivity analysis method and used as a sensitivity analysis result. Based on the sensitivity analysis result, key parameters and key parameter variation ranges are determined, and the sampling point distribution density is adjusted through an adaptive encryption strategy and simulation test cases are generated according to the sampling points. Based on the simulation test cases and a pre-set high-performance computing cluster, simulation test cases are executed in parallel, and combined with a task scheduling algorithm, simulation tests are performed with the goal of minimizing the test execution time, and the adaptability evaluation results corresponding to the drive-by-wire chassis are obtained;

[0034] A full-operating-condition adaptability evaluation index system is established, including dynamic performance index, smoothness index, energy consumption index and stability index; the performance shortcomings in the adaptability evaluation results are identified according to the full-operating-condition adaptability evaluation index system and combined with the vibration response and stress-strain distribution under different operating conditions; the association rules between different performance indicators and control parameters are identified through an association rule mining algorithm; a knowledge graph and an optimization rule base are constructed based on the association rules and performance indicators; control parameter optimization suggestions corresponding to the current performance shortcomings are generated based on the optimization rule base and returned to the drive-by-wire chassis integrated simulation model; and the drive-by-wire chassis optimization design parameters and optimal control strategy are generated by repeated iterations.

[0035] In an optional embodiment,

[0036] The road roughness is estimated based on the intensity information of point cloud reflection as shown in the following formula:

[0037]

[0038] Where R represents the comprehensive roughness estimate, S represents the total number of scales, and w s is the weight coefficient, which indicates the importance of scale s in the overall roughness calculation, N is the total number of points in the point cloud, and K s represents the number of neighborhood points at scale s, Indicates point p under scale s i The normal vector of Indicates point p under scale s j The normal vector of is the normal vector and The inner product of represents the similarity between two normal vectors.

[0039] A second aspect of an embodiment of the present invention provides a chassis control-by-wire simulation system for an intelligent networked vehicle based on multi-domain modeling, comprising:

[0040] The first unit is used to obtain structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, and at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamics equation is established, and the multi-flexible body dynamics equation is solved by a Newmark implicit and explicit algorithm to obtain vibration responses and stress-strain distributions under different working conditions;

[0041] The second unit is used to embed the multi-flexible body dynamics model into a preset drive-by-wire chassis integrated simulation model, perform multi-physics field finite element modeling through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, generate multiple finite element models, perform multi-physics field fusion at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, generate a multi-physics field model and integrate it into the drive-by-wire chassis integrated simulation model, build a multi-error source fusion model corresponding to multiple sensors, perform online error identification and active compensation through the Kalman filter algorithm, generate sensor data fusion results, feed back the sensor data fusion results, vibration responses and stress-strain distribution under different working conditions to the controller, perform closed-loop control of the drive-by-wire chassis under all working conditions in combination with the dynamic response characteristics of the drive-by-wire chassis, and generate a closed-loop integrated simulation model of the drive-by-wire chassis;

[0042] The third unit is used to collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parameterized model of the road, embed the three-dimensional parameterized model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full working condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

[0043] According to a third aspect of the embodiments of the present invention,

[0044] An electronic device is provided, comprising:

[0045] processor;

[0046] a memory for storing processor-executable instructions;

[0047] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0048] A fourth aspect of the embodiments of the present invention is:

[0049] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0050] In the present invention, by constructing an integrated simulation model of the drive-by-wire chassis including a multi-flexible body dynamics model, a multi-physics field model, and a sensor error model, and coupling it with a road parameterized model to generate a vehicle-road coupling model, the vibration response and stress-strain distribution of the drive-by-wire chassis in a complex road environment can be simulated, and its adaptability can be evaluated, providing a basis for the optimal design of the drive-by-wire chassis. By mining the drive-by-wire strategy through a machine learning method and constructing an optimization rule library, control parameter optimization suggestions can be intelligently generated to achieve self-optimization of the control strategy of the drive-by-wire chassis. The intelligent optimization design system of the drive-by-wire chassis lays a foundation for the engineering application of the drive-by-wire chassis of intelligent connected vehicles, and is of great significance for promoting the development of the intelligent connected vehicle industry. In summary, the present invention proposes a series of methods such as multidisciplinary coupling simulation and data-driven optimization, realizes the integration of multidisciplinary knowledge such as mechanics, control, and computers, and well solves the difficult problems of modeling, simulation, and optimization in the design of complex systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram of a flow chart of a method for simulating chassis control-by-wire of an intelligent networked vehicle based on multi-domain modeling according to an embodiment of the present invention;

[0052] Figure 2 It is a structural schematic diagram of an intelligent connected vehicle chassis control-by-wire simulation system based on multi-domain modeling according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0055] Figure 1 FIG. 1 is a flow chart of a method for simulating chassis control-by-wire of an intelligent networked vehicle based on multi-domain modeling according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0056] S1. Obtain structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, introduce a coupling coordination equation at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model through the Lagrange multiplier method and establish a multi-flexible body dynamics equation, solve the multi-flexible body dynamics equation through the Newmark implicit explicit algorithm, and obtain the vibration response and stress-strain distribution under different working conditions;

[0057] The material nonlinearity is that the stress and strain of the material are not linearly related, which is common in plastic materials or large deformation conditions. The collective nonlinearity is not related to material properties, but is caused by geometric deformation. The parametric finite element method is to describe the nonlinear characteristics of the problem by parameterization when dealing with nonlinear problems, and use the finite element method for discretization and solution. The finite element grid is to discretize a continuum (such as structure, material) into a finite number of nodes and units to form a grid. The multi-flexible body dynamics model is a dynamics model containing multiple flexible bodies (i.e., objects that can be deformed). The flexible body dynamics submodel is used to describe the movement and deformation of the flexible body, and is often modeled by the finite element method. The rigid body dynamics The sub-model is used to describe the motion of a rigid body, ignoring internal deformation. The coupling interface is the contact and action surface between a flexible body and a rigid body or between multiple flexible bodies. The Lagrange multiplier method is a mathematical method for dealing with constraints, which is often used in multi-body dynamics. By introducing Lagrange multipliers to impose coupling constraints, different sub-models can work in coordination. The coupling coordination equation is used to describe the coordination relationship between motion and force at the coupling interface in a multi-body system to ensure that the dynamic behaviors of different subsystems at the coupling interface are consistent. The multi-flexible body dynamics equation describes the motion equation of a system containing multiple flexible bodies, taking into account the deformation and interaction force of the flexible bodies, and is usually solved by combining finite element and multi-body dynamics methods.

[0058] In an optional embodiment,

[0059] The structural parameters and dynamic parameters of the intelligent networked vehicle chassis are obtained, a finite element mesh is constructed based on a parametric finite element method of material nonlinearity and collective nonlinearity, the finite element mesh is adaptively optimized, and a multi-flexible body dynamic model corresponding to the intelligent networked vehicle chassis is established, wherein the multi-flexible body dynamic model includes a flexible body dynamic sub-model and a rigid body dynamic sub-model, and at the coupling interface between the flexible body dynamic sub-model and the rigid body dynamic sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamic equation is established, and the multi-flexible body dynamic equation is solved by a Newmark implicit explicit algorithm, and the vibration response and stress-strain distribution under different working conditions are obtained, including:

[0060] Reconstruct the surface profile information of different components in the intelligent networked vehicle chassis through three-dimensional scanning and structured light through reverse engineering methods to obtain a three-dimensional mesh model, and obtain the structural parameters corresponding to the intelligent networked vehicle chassis based on the three-dimensional mesh model, wherein the structural parameters include the size parameters, posture parameters and connection parameters corresponding to different components, and obtain the dynamic parameters corresponding to the intelligent networked vehicle chassis, wherein the dynamic parameters include mass, stiffness, damping and moment of inertia;

[0061] Based on the structural parameters and the dynamic parameters, a finite element mesh corresponding to the chassis of the intelligent networked vehicle is constructed based on a parametric finite element method that takes into account material nonlinearity and collective nonlinearity; based on topological parameters corresponding to chassis configuration features, a logical relationship between chassis components and interfaces is determined and a consistent topological description is formed; based on the consistent topological description and the finite element mesh, corresponding gradient information of stress and strain is obtained through post-processing analysis; the gradient information is used as an error estimation indicator; the area and degree of local mesh refinement and local mesh coarsening are adaptively determined; a mesh refinement criterion and a mesh coarsening criterion are generated at the same time; and adaptive mesh optimization is performed based on the mesh refinement criterion and the mesh coarsening criterion;

[0062] A multi-flexible body dynamics model corresponding to the intelligent connected vehicle chassis is established based on the optimized finite element mesh, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics model, the flexible body dynamics sub-model determines the large deformation and large rotation corresponding to the flexible body according to the absolute node coordinate method, constructs an initial flexible body dynamics equation in combination with non-conservative forces and corresponding constraints, and performs dimensionality reduction through the modal superposition method to obtain the flexible body dynamics equation, the rigid body dynamics model is modeled through the Newton-Euler equations, defines the position vector of any point of the rigid body and determines the kinetic energy of the rigid body, establishes the translation equation and the rotation equation of the rigid body in combination with the Euler angle parameterized rotation matrix, and determines the rigid body dynamics equation based on the translation equation and the rotation equation;

[0063] Based on the rigid body dynamics equation and the flexible body dynamics equation, the coupled displacement coordination conditions and velocity coordination conditions are defined by the Lagrange multiplier method, the augmented Lagrange equation is constructed in combination with the Lagrange multiplier, the partial derivatives of the generalized coordinates are calculated and the multi-flexible body dynamics equation is constructed, the multi-flexible body dynamics equation is solved by the Newmark implicit and explicit algorithm combined with the mixed integration strategy, and the vibration response and stress-strain distribution of the connected vehicle chassis under different working conditions are obtained.

[0064] The reverse engineering method reconstructs the computational model of an existing object by measuring its geometric shape, material properties and other information, usually achieved through three-dimensional scanning and point cloud processing technology. The consistent topological description is used in reverse engineering to keep the reconstructed model consistent with the topological structure of the original object, that is, to ensure that it is consistent with the original object in terms of geometry, connectivity, etc. The mesh refinement criterion is used to determine which areas in the finite element mesh require more details, and is usually refined in areas of stress concentration or large deformation to improve calculation accuracy. The mesh coarsening criterion is opposite to the refinement criterion, and determines which areas can be coarsened using fewer units, thereby reducing the amount of calculation and improving calculation efficiency. The absolute node coordinate method is a modeling method used in multi-body systems. The movement of the nodes is described by an absolute coordinate system, which is particularly suitable for dealing with dynamic problems of flexible bodies. The non-conservative force refers to a force that does not meet the conservative force conditions, such as friction, damping, etc. The modal superposition method refers to the dynamic response of complex structures. It should be decomposed into the superposition of its various modes. The Newton-Euler equation is an equation used to describe the motion of a rigid body. It combines Newton's second law and Euler's law of translation and rotation to calculate the acceleration, velocity and position of a rigid body. The Euler angle parameterized rotation matrix is ​​used to describe the rotation as a successive rotation around three rotation axes of a fixed coordinate system. It is usually used as a rotation representation method in rigid body dynamics. The translation equation is used to describe the translational motion of an object under the action of an external force. The rotation equation is used to describe the rotational motion of an object around its center of mass or a fixed axis. The rotational dynamics of a rigid body are described by a rotation matrix and angular momentum. The augmented Lagrangian equation is an equation used to deal with the dynamics of a multi-body system with constraints. By introducing augmented Lagrangian multipliers, it is ensured that the constraints of the system are met. The hybrid integration strategy combines implicit and explicit time integration methods, which not only maintains the efficiency of the explicit method, but also utilizes the stability of the implicit method. It is suitable for the dynamics solution of multi-flexible body systems.

[0065] Use a 3D scanner to perform an all-round scan of each component of the intelligent networked vehicle chassis to obtain high-precision point cloud data. The scanning process requires reasonable setting of scanning parameters, such as scanning distance, scanning angle, scanning accuracy, etc., to ensure that complete and accurate point cloud data is obtained. Use structured light technology to perform high-precision scanning on the surface of key chassis components. Structured light scanning projects specific grating stripes onto the surface of the object to be measured, captures the stripe image through a camera, and combines the principle of triangulation to calculate the 3D coordinates of each point on the surface to obtain high-precision surface profile information. The obtained point cloud data and surface profile information are imported into the reverse engineering software. Through denoising, registration, fusion and other operations, the 3D CAD model of the intelligent networked vehicle chassis is reconstructed. The 3D reconstruction process needs to consider the geometric features and topological relationships of different components to ensure that the reconstructed 3D model is consistent with the actual chassis structure. The reconstructed 3D CAD model is meshed to generate a 3D mesh model suitable for finite element analysis and multi-body dynamics analysis.

[0066] Based on the three-dimensional mesh model, the structural parameters of the intelligent connected vehicle chassis are extracted, including the size parameters (length, width, height, etc.), posture parameters (position, posture, etc.) and connection parameters (connection method, connection position, etc.) of different components. According to the material properties and structural characteristics of each chassis component, the corresponding dynamic parameters are obtained, including mass, stiffness, damping and moment of inertia. The mass parameter can be calculated through the volume and material density of the three-dimensional model, the stiffness parameter can be estimated through the elastic modulus of the material and the geometric characteristics of the structure, the damping parameter can be determined according to the damping characteristics of the material and the energy dissipation mechanism of the structure, and the moment of inertia can be calculated based on the shape and mass distribution of the three-dimensional model.

[0067] Based on the obtained structural parameters and dynamic parameters, a parametric finite element method that considers material nonlinearity and geometric nonlinearity is used to construct a finite element model of the chassis of an intelligent networked vehicle. According to the chassis configuration characteristics, the topological relationship between chassis components and between components and interfaces is determined to form a consistent topological description. The consistent topological description needs to consider the geometric shapes, connection methods and assembly relationships of different components to ensure that the finite element model can accurately reflect the structural characteristics and force transmission mechanism of the chassis. Through finite element calculation, the stress and strain distribution of the chassis under different working conditions is obtained. The gradient information of stress and strain is extracted using post-processing technology as the error estimation index for mesh optimization. According to the error estimation index, the areas that need mesh refinement and mesh coarsening are adaptively determined, and the degree of refinement and coarsening is quantified. At the same time, the corresponding mesh refinement criteria and mesh coarsening criteria are generated to guide the mesh optimization process. The adaptive mesh optimization method is used to reduce the number of meshes and improve the computational efficiency while ensuring the computational accuracy.

[0068] For flexible components in the chassis, such as control arms and shock absorbers, the absolute node coordinate method (ANCF) is used to establish a flexible body dynamics model. In order to improve the calculation efficiency, the modal superposition method is used to reduce the dimension of the flexible body dynamics equation to obtain the flexible body dynamics equation after dimension reduction. For rigid components in the chassis, such as the frame and wheel hub, the Newton-Euler equation is used for modeling, the position vector of any point on the rigid body is defined, the kinetic energy of the rigid body is calculated, and the rotation matrix parameterized by the Euler angle is used to establish the translation equation and rotation equation of the rigid body to obtain the rigid body dynamics equation. The flexible body dynamics equation and the rigid body dynamics equation are coupled to construct an intelligent networked vehicle. The multi-flexible body dynamics model of the chassis adopts the Lagrange multiplier method, introduces the coupled displacement coordination condition and the velocity coordination condition, constructs the augmented Lagrange equation, obtains the partial derivative of the generalized coordinates, and obtains the dynamic equation of the multi-flexible body system. The Newmark implicit-explicit hybrid integration algorithm is used to solve the multi-flexible body dynamics equation of the intelligent networked vehicle chassis. According to the actual working conditions of the chassis, such as bumpy roads, emergency braking, etc., reasonable initial conditions and boundary conditions are set. Through numerical integration, the vibration response and stress-strain distribution of each chassis component under different working conditions are obtained, providing data support for the optimization design and control strategy formulation of the chassis;

[0069] Exemplarily, a handheld 3D scanner and a structured light scanning system are used to perform 3D scanning on key components of a McPherson front suspension, such as a frame, a lower control arm, a steering knuckle, a spring, and a shock absorber, to obtain point cloud data and surface profile information, and the obtained 3D data is processed using reverse engineering software, and a 3D CAD model of the McPherson front suspension is reconstructed through operations such as denoising, registration, and subdivision, and a 3D mesh model suitable for finite element analysis is generated, and structural parameters of the McPherson front suspension are extracted from the 3D mesh model, including the length, width, and thickness of the lower control arm, and the installation position and installation angle of the spring and shock absorber, etc. According to the material properties of the components, the corresponding dynamic parameters such as mass, stiffness, damping, and moment of inertia are obtained, and a parametric finite element model of the McPherson front suspension is constructed in the finite element analysis software, and material properties, boundary conditions, and contact relationships are defined, and the natural frequency, vibration mode, and stress-strain distribution of the suspension are obtained through modal analysis and static analysis;

[0070] By using the post-processing function of the finite element analysis software, the gradient information of stress and strain is extracted as the error estimation index for mesh adaptive optimization. According to the error level, the mesh refinement area (such as the stress concentration area) and the mesh coarsening area (such as the stress gradient flat area) are determined, and the mesh refinement and coarsening criteria are generated. The finite element model is adaptively optimized to obtain the optimized finite element model. The multi-flexible body dynamics model of the McPherson front suspension is constructed in the multi-body dynamics simulation software, and the flexible body dynamics model of the lower control arm and spring is established by the absolute node coordinate method. The frame, steering knuckle, etc. are established using the Newton-Euler equation. The dynamic model of the rigid body is introduced through the Lagrange multiplier method, and the coupled displacement coordination conditions and velocity coordination conditions of the McPherson front suspension are introduced to construct the multi-flexible body dynamic equation of the suspension system. For bumpy road conditions, the road spectrum input is set, and the multi-flexible body dynamic response of the McPherson front suspension is solved using the numerical integration algorithm to obtain the vibration response and stress-strain distribution of the suspension components. Based on the obtained vibration response and stress-strain results, the dynamic performance of the McPherson front suspension is evaluated, and the suspension parameters are optimized, such as adjusting the spring stiffness, shock absorber damping, etc., to improve the suspension smoothness, handling stability and ride comfort.

[0071] In this embodiment, through three-dimensional scanning and reverse engineering technology, the geometric morphology and structural parameters of the chassis of the intelligent connected vehicle are obtained efficiently and accurately, laying a data foundation for subsequent modeling and analysis. Through parametric modeling technology, the rapid update and iteration of the chassis finite element model is realized, and the modeling efficiency is improved. The adaptive grid optimization method can automatically adjust the grid density according to the stress-strain gradient, realize grid encryption in the stress concentration area, and realize grid coarsening in the stress gradient area. Under the premise of ensuring the calculation accuracy, the calculation scale and calculation cost are greatly reduced. The absolute node coordinate method is used to describe the large deformation and large rotation of the flexible component, and the modal superposition technology is combined to achieve dimensionality reduction. It not only takes into account the flexibility effect, but also ensures the calculation efficiency, making the dynamic analysis results more reliable. In summary, this embodiment realizes the digitization and intelligence of the entire process of the chassis from geometric modeling to performance analysis, greatly improves the efficiency and quality of chassis research and development, and lays a solid theoretical and technical foundation for the optimization design and control strategy formulation of the intelligent connected vehicle chassis, with significant technical innovation and engineering application value.

[0072] In an optional embodiment,

[0073] The multi-flexible body dynamics equation is solved by using the Newmark implicit and explicit algorithm combined with the mixed integration strategy as shown in the following formula:

[0074]

[0075] Among them, q n+1 represents the generalized coordinate vector of the next time step n+1, q nRepresents the generalized coordinate vector of the current time step, △t n Represents the adaptive step size of the current time step, q′ n Represents the velocity vector of the current time step, q″ n represents the acceleration vector of the current time step, q″(n+1) represents the acceleration vector of the next time step, and β represents the acceleration control parameter.

[0076] In this embodiment, a hybrid integration strategy is used to maximize the computational efficiency and shorten the dynamics simulation time while ensuring computational accuracy and stability. Adaptive time step control is used to optimize the computational efficiency and accelerate the dynamics simulation speed while ensuring computational accuracy and stability. Compared with traditional single integration strategies and fixed-step integration, it can better balance computational accuracy, computational efficiency and numerical stability, and provide an efficient and reliable numerical simulation tool for the dynamics performance analysis and optimization design of the chassis of intelligent connected vehicles. In summary, this embodiment forms a set of efficient, accurate and stable dynamics numerical solution solutions, which lays a solid theoretical foundation and methodological support for the dynamics modeling, performance analysis and optimization design of the chassis of intelligent connected vehicles.

[0077] S2. embed the multi-flexible body dynamics model into a preset drive-by-wire chassis integrated simulation model, perform multi-physics field finite element modeling through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, generate multiple finite element models, perform multi-physics field fusion at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, generate a multi-physics field model and integrate it into the drive-by-wire chassis integrated simulation model, build a multi-error source fusion model corresponding to multiple sensors, perform online error identification and active compensation through the Kalman filter algorithm, generate sensor data fusion results, feed back the sensor data fusion results, vibration response and stress-strain distribution under different working conditions to the controller, perform closed-loop control of the drive-by-wire chassis under all working conditions in combination with the dynamic response characteristics of the drive-by-wire chassis, and generate a closed-loop integrated simulation model of the drive-by-wire chassis;

[0078] The drive-by-wire chassis integrated simulation model is used to simulate and verify the overall performance of the drive-by-wire chassis system, including aspects such as handling, stability, and comfort, and integrates various subsystems and their dynamic responses. The drive-by-wire actuator subsystem is responsible for executing the vehicle's control commands, and is usually composed of an electric motor, a sensor, and a feedback system, which directly affects the vehicle's handling performance and response. The multi-physics finite element modeling combines multiple physical fields (such as mechanics, fluid, thermal, etc.) for modeling, and uses the finite element method to simulate complex engineering problems. It is particularly suitable for systems that need to consider multiple physical effects. The finite element model divides the structure or material into discrete finite elements, establishes a mathematical model to solve complex physical problems, and can simulate the behavior of the structure under different loads and boundary conditions. The fluid-solid coupling method is used to simultaneously consider the interaction between fluid and solid, and is suitable for applications involving fluid flow and solid deformation, such as aerodynamics and body structure in vehicle dynamics. The multi-physics fusion combines multiple physical fields The model is integrated into a unified computing framework so that the mutual influence of each physical field can be considered simultaneously during the analysis. The multi-physics field model describes the model of multiple physical phenomena in a system, such as considering mechanical, fluid and thermal effects simultaneously in vehicle dynamics. The multi-error source fusion model is used to integrate errors from multiple sensors or computing models to improve the overall accuracy and reliability of the system. The online error identification is real-time monitoring and identification of error sources in the system, usually combining sensor data and algorithms to automatically adjust the control strategy. The drive-by-wire chassis dynamic response characteristics describe the dynamic behavior of the drive-by-wire chassis under different working conditions, including response speed, stability and handling performance. The full-condition drive-by-wire chassis closed-loop control adjusts the control strategy in real time under various working conditions to ensure the stability and safety of the drive-by-wire chassis system in complex environments. The drive-by-wire chassis closed-loop integrated simulation model combines the closed-loop control algorithm with the drive-by-wire chassis model to simulate the performance of the control system in a real environment for optimization and verification.

[0079] In an optional embodiment,

[0080] The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, and multiple finite element models are generated. Multi-physics field fusion is performed at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, and a multi-physics field model is generated and integrated into the drive-by-wire chassis integrated simulation model. A multi-error source fusion model corresponding to multiple sensors is constructed, and online error identification and active compensation are performed through the Kalman filter algorithm to generate sensor data fusion results. The sensor data fusion results, vibration responses and stress-strain distributions under different working conditions are fed back to the controller, and closed-loop control of the drive-by-wire chassis under all working conditions is performed in combination with the dynamic response characteristics of the drive-by-wire chassis. The generation of a closed-loop integrated simulation model of the drive-by-wire chassis includes:

[0081] The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, an electromagnetic field finite element model is constructed by combining the Maxwell equations and the geometric structure of the motor, the electromagnetic field equations are solved by a meshing method to obtain the magnetic field distribution data inside the motor, a fluid field finite element model is constructed according to the fluid flow and fluid control equations inside the hydraulic cylinder, and the fluid field equations are solved to obtain the velocity field and pressure field corresponding to the fluid, solid mechanics control equations are selected according to the solid structure part of the actuator and a solid field finite element model is constructed, and the solid field equations are solved in combination with the geometric shape and load conditions of the solid structure part to obtain the displacement field and stress field of the solid structure part;

[0082] Based on the electromagnetic field finite element model, the fluid field finite element model and the solid field finite element model, a coupling interface between an actuator and a transmission mechanism is determined, coupling conditions are defined at the coupling interface, a mathematical constraint relationship of the coupling interface is established by a penalty function method, each finite element model is solved at each time step based on the mathematical constraint relationship, multi-physics field fusion is achieved by information exchange and iterative correction at the coupling interface, and the multi-physics field model is obtained and integrated into the integrated simulation model of the drive-by-wire chassis;

[0083] A corresponding measurement model is constructed based on the preset sensors, the sensor error sources are determined based on the measurement model, and the multi-error source fusion model is constructed through the Kalman filter algorithm, the measurement data of each sensor is acquired in real time, and online error identification and active compensation are performed according to the multi-error source fusion model, the data after active compensation are used as the sensor data fusion result, combined with the vibration response and stress-strain distribution feedback and the dynamic response characteristics of the drive-by-wire chassis under the different working conditions, the closed-loop control of the drive-by-wire chassis under all working conditions is performed through a preset controller, and the closed-loop integrated simulation model of the drive-by-wire chassis is obtained.

[0084] The grid division method is a strategy for dividing the computational domain into finite elements, which usually includes uniform grids, non-uniform grids, adaptive grids, etc., to ensure computational accuracy and efficiency. The fluid control equations describe the basic equations of fluid motion, such as the Navier-Stokes equations, which are used to simulate fluid flow and interaction. The solid structure part describes the mechanical behavior of solid materials, and elasticity or plasticity theory is usually used to establish the corresponding physical model. The coupling conditions describe the conditions for interaction between fluid and solid, such as boundary conditions and the transmission relationship of interaction forces. The penalty function method is a method for processing constraint conditions. The mathematical constraint relationship is a constraint condition defined in modeling, which is usually used to ensure the physical rationality of the model and the stability of the solution process.

[0085] In the wire-controlled chassis integrated simulation software, the system architecture of the wire-controlled chassis is built, including subsystems such as actuators, transmission mechanisms, controllers, sensors, etc. The established multi-flexible body dynamics model is imported into the wire-controlled chassis integrated simulation model as the dynamics model of the wire-controlled chassis. In the wire-controlled chassis integrated simulation model, the interface and connection relationship between the multi-flexible body dynamics model and other subsystems are defined. According to the motor geometry of the actuator, the electromagnetic field finite element method is used, combined with Maxwell's equations, to build the electromagnetic field finite element model of the motor, mesh the electromagnetic field finite element model, and solve the electric Magnetic field equations are used to obtain the magnetic field distribution data inside the motor, such as magnetic induction intensity, magnetic flux density, etc. According to the flow characteristics of the fluid inside the hydraulic cylinder of the actuator, the fluid mechanics finite element method is used to construct a fluid field finite element model of the hydraulic cylinder, and the fluid field equations are solved to obtain the velocity field and pressure field distribution of the fluid inside the hydraulic cylinder. According to the solid structural parts of the actuator, such as the piston rod and cylinder body, the applicable solid mechanics control equations are selected to construct a solid field finite element model. In combination with the geometric shape and load conditions of the solid structural parts, the solid field equations are solved to obtain the displacement field and stress field distribution of the solid structure;

[0086] Based on the constructed finite element models of electromagnetic field, fluid field and solid field, the coupling interface between the physical fields inside the actuator is determined, such as the coupling interface between the electromagnetic field and the fluid field, the coupling interface between the fluid field and the solid field, etc. At the coupling interface, the coupling conditions between different physical fields are defined, such as electromagnetic force, fluid force, solid stress, etc. The penalty function method is used to convert the coupling conditions at the coupling interface into mathematical constraint relationships, and a mathematical description of the coupling interface is established. Based on the mathematical constraint relationships of the coupling interface, the finite element model of each physical field is solved at each time step. During the solution process of each time step, information exchange and iterative correction are performed between the physical fields at the coupling interface to realize the interaction and comprehensive analysis of multiple physical fields, complete the construction of the multi-physical field coupling model, and integrate it into the integrated simulation model of the drive-by-wire chassis.

[0087] According to the types and characteristics of sensors used in the drive-by-wire chassis, such as position sensors, speed sensors, force sensors, etc., a corresponding sensor measurement model is constructed. In the sensor measurement model, various error sources, such as noise, drift, nonlinearity, etc., are considered to establish the relationship between the error source and the measurement error. The Kalman filter algorithm is used to fuse multiple error sources of each sensor to construct a multi-error source fusion model. The measurement data of each sensor is acquired in real time. Online error identification and active compensation are performed according to the multi-error source fusion model. The sensor data after active compensation is used as the result of sensor data fusion for state feedback and closed-loop control of the drive-by-wire chassis. In the real model, different working conditions are set, such as straight-line driving, steering, braking, etc., and the integrated simulation model is run to obtain the dynamic response characteristics such as vibration response and stress-strain distribution of the drive-by-wire chassis under different working conditions. The dynamic response characteristics are compared and fed back with the sensor data fusion results to evaluate the real-time status and performance of the drive-by-wire chassis. According to the pre-designed control algorithms and controllers, such as PID control, sliding mode control, and robust control, the input signal of the drive-by-wire actuator is calculated and adjusted, and the control signal output by the controller is applied to the drive-by-wire actuator to achieve closed-loop control of the drive-by-wire chassis under all working conditions, and iterate and optimize continuously until satisfactory control performance and system response are achieved;

[0088] Integrate multi-physics field models, sensor measurement models, multi-error source fusion models, and controllers into the integrated simulation model of the drive-by-wire chassis, test and verify the integrated simulation model, evaluate its accuracy, stability, and real-time performance, and make necessary corrections and optimizations to the integrated simulation model based on the test results to improve its simulation performance and reliability. Complete the construction of the closed-loop integrated simulation model of the drive-by-wire chassis, and provide an efficient and accurate simulation tool for the design, optimization, and control of the drive-by-wire chassis.

[0089] For example, taking the wire-controlled steering system of an electric vehicle as an example, a multi-flexible body dynamics model of the wire-controlled steering system is constructed, including components such as the steering gear, steering column, and rack. The kinematic constraints and force transmission relationships between the components are considered, and the multi-flexible body dynamics model is embedded in the integrated simulation model of the wire-controlled steering system. Taking the steering gear motor as an example, the motor stator winding adopts a 12-slot, 10-pole structure, and the permanent magnet is made of neodymium iron boron material. The motor is modeled using electromagnetic field finite element software to obtain the electromagnetic torque of the motor under different excitation currents. The hydraulic cylinder inside the steering gear adopts a double-acting structure. Through the fluid mechanics finite element method, the motor is modeled using electromagnetic field finite element software. The software is used to model the steering gear and obtain the pressure distribution inside the hydraulic cylinder under different steering torques. The steering gear housing is made of aluminum alloy. The solid mechanics finite element software is used to perform stress analysis on it and obtain the stress distribution under the steering torque. The coupling conditions of electromagnetic force, hydraulic pressure and solid stress are defined between the finite element models of the steering gear motor, hydraulic cylinder and housing. The penalty function method is used to establish the mathematical constraint relationship of the coupling interface. At each simulation time step, the electromagnetic field, fluid field and solid field equations are solved simultaneously, and the physical quantities are exchanged and iterated at the coupling interface to obtain the comprehensive response of the steering gear under the coupling of multiple physical fields.

[0090] According to the characteristic parameters of the angle sensor and torque sensor used in the steering system, the angle error model and torque error model are established respectively. The error model includes zero position error, sensitivity error, nonlinear error, etc. The error models of the two sensors are fused using the Kalman filter algorithm to obtain a multi-error source fusion model of angle and torque. During the steering process of the actual vehicle, sensor data is collected in real time, and error compensation is performed through the multi-error source fusion model. The compensated angle and torque are used as feedback values. Multiple steering conditions are set, such as straight-line driving steering, low-speed steering, high-speed steering, snaking conditions, etc. The integrated simulation model of the wire-controlled steering system is run to obtain the vibration response and stress distribution of the steering gear under different conditions. The vibration and stress signals of the steering gear are closed-loop fed back with the fusion values ​​of the angle and torque sensors. According to the designed fuzzy PID controller, the control of the steering motor is calculated to achieve closed-loop control of the steering gear under all working conditions.

[0091] The steering gear multi-physics field model, angle torque sensor measurement model, multi-error source fusion model and fuzzy PID controller are integrated into the closed-loop integrated simulation model of the steering-by-wire system. The integrated model is calibrated to obtain the damping characteristics and self-centering characteristics of the steering gear. The results are compared with the bench test data to verify the accuracy of the integrated simulation model. Finally, a complete closed-loop integrated simulation model of the steering-by-wire system is obtained, which is used for the forward design and iterative optimization of the steering system.

[0092] In this embodiment, by embedding the multi-flexible body dynamics model into the integrated simulation model of the drive-by-wire chassis, the coordinated simulation of the dynamic characteristics of the drive-by-wire chassis system and other subsystems is achieved, the accuracy and reliability of the integrated simulation of the drive-by-wire chassis are improved, and an efficient and accurate simulation tool is provided for the design and optimization of the drive-by-wire chassis. The multi-physics finite element modeling method is used to conduct a comprehensive and in-depth analysis of the drive-by-wire actuator, and the interaction and coupling effects of the electromagnetic field, fluid field and solid field are considered. The distribution laws of the magnetic field, velocity field, pressure field, displacement field and stress field inside the actuator are obtained, which provides an important basis for the design and optimization of the actuator. The surface and mathematical constraint relationship are established, and the multi-physical field fusion of electromagnetic field, fluid field and solid field is realized. It can accurately describe the multi-physical field interaction and energy transfer process inside the actuator, improve the physical reality and numerical convergence of the integrated simulation of the drive-by-wire chassis, construct a sensor measurement model and a multi-error source fusion model, realize the online identification and active compensation of sensor errors, effectively improve the accuracy and reliability of sensor data, and lay the foundation for the state feedback and closed-loop control of the drive-by-wire chassis. In summary, this embodiment provides a new idea and method for the forward design, multidisciplinary optimization and closed-loop control of the drive-by-wire chassis, which is of great significance to improving the performance indicators of the drive-by-wire chassis and promoting the industrial application of drive-by-wire technology.

[0093] In an optional embodiment,

[0094] The mathematical constraint relationship of the coupling interface is established by the penalty function method, and each finite element model is solved at each time step based on the mathematical constraint relationship as shown in the following formula:

[0095]

[0096] Among them, ε total is the overall energy function value, which represents the sum of the finite element model energy and penalty function, ε A is the energy of finite element model A, ε B represents the energy of the finite element model B, k i represents the penalty factor at level i, u A represents the displacement vector of the finite element model A on the coupling interface, u B represents the displacement vector of the finite element model B on the coupling interface, is an indicator function, indicating that when the displacement difference on the coupling interface is greater than the threshold ζ i 1 when ζ is set, otherwise 0. i Represents the displacement difference threshold of the i-th level.

[0097] In this embodiment, by introducing a penalty function at the coupling interface, the finite element models of different physical fields are coupled through the displacement continuity condition, thereby realizing the interaction and energy transfer between the electromagnetic field, the fluid field and the solid field, and improving the accuracy and reliability of the multi-physical field coupling analysis. The penalty function method introduces a penalty term in the energy function to constrain and penalize the displacement difference on the coupling interface, so that in the iterative solution process, the displacement difference on the coupling interface gradually decreases to meet the displacement continuity condition, thereby improving the stability and convergence speed of the numerical calculation. By introducing a graded penalty factor and a displacement difference threshold, adaptive adjustment of the coupling interface constraint conditions is achieved. By adjusting the indicator function and the threshold in the penalty function, dynamic control of the coupling interface constraint conditions can be achieved to adapt to changes in the multi-physical field coupling problem. In summary, this embodiment provides reference and inspiration for the modeling and solution of other multi-physical field coupling problems (such as fluid-solid coupling, thermoelectric coupling, electrochemical coupling, etc.), and promotes the development of multi-physical field coupling theory and methods.

[0098] S3. Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, pavement attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parametric model of the road, embed the three-dimensional parametric model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full-condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and iterate repeatedly to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

[0099] The multi-source heterogeneous road map data refers to road information from different sensors and data sources. The vehicle-road coupling model describes the interaction between vehicles and roads, including vehicle dynamics, road characteristics and traffic flow, aiming to optimize traffic management and improve driving safety. The intelligent sampling refers to the dynamic selection of sample points based on data characteristics and needs to improve the representativeness and processing efficiency of the data, and is often used in the analysis and modeling of large-scale data sets. The adaptive encryption method involves dynamically adjusting the encryption strategy according to data traffic and security requirements to ensure the security and real-time nature of data transmission, which is especially important in vehicle communications. The full-condition evaluation refers to a comprehensive evaluation of system performance under various working conditions to ensure the reliability and stability of the system under different environments and loads. The wire control strategy involves managing various vehicle functions such as steering, braking and acceleration through electronic control to achieve higher accuracy and response speed, and improve the vehicle's handling and safety.

[0100] In an optional embodiment,

[0101] Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parametric model of the road, embed the three-dimensional parametric model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full-condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis, including:

[0102] Collecting multi-source heterogeneous road map data corresponding to different cities, wherein the multi-source heterogeneous road map data includes high-precision maps, laser point clouds and street view images, obtaining the spatiotemporal references of different data sources based on inertial navigation technology and performing point cloud registration through control points, generating consistent point cloud data, and preprocessing the consistent point cloud data through a voxel filtering algorithm to obtain preprocessed multi-source heterogeneous road map data;

[0103] For the pre-processed multi-source heterogeneous road map data, an end-to-end segmentation model is selected to construct a labeled data set for road semantic segmentation, wherein the labeled data set includes road images under different weather and lighting conditions, pixel-level annotation truth values ​​are generated by combining manual annotation and automatic generation, training is performed in combination with a cross entropy loss function and the end-to-end segmentation model is optimized by an optimizer, and a performance evaluation is performed using a pre-set independent test set. If the end-to-end segmentation model meets the pre-set performance standard, the end-to-end segmentation model is used to perform semantic segmentation on the processed multi-source heterogeneous road map data, identify urban roads, and extract road center lines and boundary lines as geometric features through a feature extraction algorithm, estimate road surface roughness and road surface material based on the intensity information of point cloud reflection and use them as road surface attribute parameters, extract environmental information corresponding to different roads based on the street view images, and extract environmental parameters through a feature extraction algorithm, and construct a three-dimensional parameterized road model based on the geometric features, road surface attribute parameters, and environmental parameters, and embed the three-dimensional parameterized road model into the closed-loop integrated simulation model of the drive-by-wire chassis to generate a vehicle-road coupling model;

[0104] Based on the vehicle-road coupling model, the high-dimensional parameter space corresponding to the three-dimensional parameterized model of the road is explored through an intelligent sampling method, and the influence of different dimensional parameters on vehicle performance is evaluated in combination with a sensitivity analysis method and used as a sensitivity analysis result. Based on the sensitivity analysis result, key parameters and key parameter variation ranges are determined, and the sampling point distribution density is adjusted through an adaptive encryption strategy and simulation test cases are generated according to the sampling points. Based on the simulation test cases and a pre-set high-performance computing cluster, simulation test cases are executed in parallel, and combined with a task scheduling algorithm, simulation tests are performed with the goal of minimizing the test execution time, and the adaptability evaluation results corresponding to the drive-by-wire chassis are obtained;

[0105] A full-operating-condition adaptability evaluation index system is established, including dynamic performance index, smoothness index, energy consumption index and stability index; the performance shortcomings in the adaptability evaluation results are identified according to the full-operating-condition adaptability evaluation index system and combined with the vibration response and stress-strain distribution under different operating conditions; the association rules between different performance indicators and control parameters are identified through an association rule mining algorithm; a knowledge graph and an optimization rule base are constructed based on the association rules and performance indicators; control parameter optimization suggestions corresponding to the current performance shortcomings are generated based on the optimization rule base and returned to the drive-by-wire chassis integrated simulation model; and the drive-by-wire chassis optimization design parameters and optimal control strategy are generated by repeated iterations.

[0106] The inertial navigation technology is used to determine the current position and motion state of an object, usually relying on accelerometers and gyroscopes, and is widely used in fields such as autonomous driving and drones. The point cloud registration is the process of aligning point cloud data from different sources or times to facilitate further analysis and modeling. The voxel filtering algorithm processes the three-dimensional point cloud data by dividing it into voxels (3D pixels), thereby effectively reducing data noise and improving the quality and accuracy of the point cloud. The pixel-level annotation truth is the process of classifying and annotating each pixel in an image or video, and is often used to train and verify deep learning models, especially in the field of autonomous driving. The three-dimensional parameterized model of the road is a mathematical model that describes the geometric characteristics and attributes of the road, and is used to support the simulation and analysis of road traffic flow and its interaction with vehicles. The sensitivity analysis method is used to evaluate the response of the model to changes in input parameters to identify key factors and optimize the design and decision-making process, and is often used in the performance analysis of complex systems. The simulation test case is a specific scenario and condition designed to verify the performance of the system or model, and is used for system testing and evaluation of its performance and stability under various conditions.

[0107] Collect multi-source heterogeneous road map data from different cities, including high-precision maps, laser point clouds, and street view images. Use inertial navigation technology to obtain the spatiotemporal benchmarks of different data sources, and perform point cloud registration through control points to generate consistent point cloud data. Use a speed-up filtering algorithm to preprocess the consistent point cloud data to obtain preprocessed multi-source heterogeneous road map data.

[0108] Select an end-to-end segmentation model, build an annotated dataset for road semantic segmentation, including road images under different weather and lighting conditions, generate pixel-level annotation truth values ​​by combining manual annotation and automatic generation, train the end-to-end segmentation model with the cross entropy loss function, and use the optimizer to optimize the model. Use a pre-set independent test set to evaluate the model's performance. If the preset performance standards are met, use the model to perform semantic segmentation on the pre-processed multi-source heterogeneous road map data.

[0109] Using the semantic segmentation results, urban roads are identified, and the center lines and boundary lines of the roads are extracted as geometric features using a feature extraction algorithm. The road roughness and road material are estimated based on the intensity information of the point cloud reflection as road attribute parameters. Environmental information corresponding to different roads is extracted based on street view images, and environmental parameters are extracted using a feature extraction algorithm. A three-dimensional parametric model of the road is constructed based on geometric features, road attribute parameters, and environmental parameters. The three-dimensional parametric model of the road is embedded in the closed-loop integrated simulation model of the drive-by-wire chassis to generate a vehicle-road coupling model.

[0110] An intelligent sampling method is used to explore the high-dimensional parameter space corresponding to the three-dimensional parametric model of the road. The sensitivity analysis method is combined to evaluate the influence of different dimensional parameters on vehicle performance. Based on the sensitivity analysis results, key parameters and key parameter variation ranges are determined. The sampling point distribution density is adjusted through an adaptive encryption strategy, and simulation test cases are generated according to the sampling points. Based on the simulation test cases and the pre-set high-performance computing cluster, the simulation test cases are executed in parallel. Combined with the task scheduling algorithm, simulation tests are performed with the goal of minimizing the test execution time, and the adaptability evaluation results corresponding to the drive-by-wire chassis are obtained.

[0111] Establish an all-condition adaptability evaluation index system, including dynamic index, smoothness index, energy consumption index and stability index. According to the all-condition adaptability evaluation index system, identify the performance shortcomings in the adaptability evaluation results. Combined with the vibration response and stress-strain distribution under different working conditions, identify the association rules between different performance indicators and control parameters through the association rule mining algorithm. Based on the association rules and performance indicators, build a knowledge graph and optimization rule base. Generate control parameter optimization suggestions corresponding to the current performance shortcomings based on the optimization rule base, and return them to the integrated simulation model of the drive-by-wire chassis. Repeat the iteration to generate the optimized design parameters and optimal control strategy of the drive-by-wire chassis.

[0112] For example, a certain automobile company develops a wire-controlled chassis. In order to evaluate its adaptability in different urban road environments, it collects high-precision maps, laser point clouds and street view image data from three cities, uses inertial navigation technology to obtain the spatiotemporal benchmarks of different data sources, and uses 500 control points to perform point cloud registration to generate consistent point cloud data. The consistent point cloud data is preprocessed using a voxel filtering algorithm to obtain preprocessed multi-source heterogeneous road map data. The end-to-end segmentation model DeepLabV3+ based on convolutional neural networks is selected to construct a labeled dataset containing 10,000 road images under different weather conditions (sunny, cloudy, rainy) and lighting conditions (daytime, nighttime). The pixel-level annotation truth is generated by combining manual annotation and automatic generation. The model is trained and optimized using the cross entropy loss function and the Adam optimizer. After 100 epochs of training, The performance was evaluated on 2,000 independent test sets, with a pixel accuracy of 95.2%, meeting the preset performance standards. The trained segmentation model was used to perform semantic segmentation on the pre-processed multi-source heterogeneous road map data, identify urban roads, extract road edges using the Canny operator, and extract road centerlines using the skeleton extraction algorithm as the geometric features of the road. The road surface roughness was divided into five levels based on the intensity information of the LiDAR point cloud reflection, and the road surface material was divided into three categories: asphalt, cement, and gravel. Based on street view images, environmental information such as building density and greening rate within 5-100 meters on both sides of the road was extracted, and the principal component analysis method was used to reduce the dimension to obtain 5-dimensional environmental parameters. Based on the geometric features, road surface attribute parameters, and environmental parameters, a three-dimensional parametric model of the road was constructed and embedded in the CarSim-Matlab joint simulation model of the drive-by-wire chassis to generate a vehicle-road coupling model.

[0113] In the vehicle-road coupling simulation test, the Latin hypercube sampling method was used to sample the 20-dimensional parameter space of the three-dimensional parametric model of the road 1,000 times. The Sobol sensitivity analysis method was used to evaluate the influence of different parameters on six performance indicators such as vehicle lateral acceleration, longitudinal acceleration, and suspension travel. The sampling point density in the key parameter area was increased to 5 times through an adaptive encryption strategy, and a total of 5,000 simulation test cases were generated. A high-performance computing cluster with 500 CPU cores was used in combination with the minimum heap scheduling algorithm to perform parallel simulation execution on 5,000 test cases. The total calculation time was 2.5 hours. In the adaptability evaluation, an evaluation system including 15 indicators such as maximum vehicle speed, acceleration time (0-100km / h), fuel consumption per 100 kilometers, and maximum yaw angular velocity was established. The simulation results were statistically analyzed and it was found that the yaw stability of the drive-by-wire chassis was insufficient under mountain roads (road curvature greater than 0.005m^-1, slope greater than 5%), and the maximum yaw angular velocity exceeded 5deg / s. The Apriori association rule mining algorithm was used to find that the support of the three control parameters, namely yaw stability, steering response time and steering angular velocity, was greater than 60%, thereby generating three optimization control rules. Based on the rule base, optimization suggestions such as "increasing the yaw angular velocity feedback gain by 20%" were generated and returned to the vehicle-road coupling model for verification. After three iterative optimizations, the yaw stability of the drive-by-wire chassis was improved, and the maximum yaw angular velocity was reduced to below 3.2deg / s, meeting the design requirements. The 30 optimization design parameters of the drive-by-wire chassis and the optimal control strategy containing 15 parameters were output.

[0114] In this embodiment, the simulation test can fully consider the vehicle-road interaction, improve the authenticity and reliability of the adaptability evaluation of the drive-by-wire chassis, and utilize high-performance computing clusters and parallel task scheduling algorithms to achieve rapid execution of large-scale simulation test cases, greatly shortening the evaluation time. An evaluation system with 15 indicators is constructed from multiple dimensions such as power, smoothness, energy consumption, and stability. It can comprehensively judge the performance of the drive-by-wire chassis under different working conditions, and provide a quantitative basis for the optimal design of the drive-by-wire chassis. Optimization suggestions are automatically generated based on the rule base and fed back to the vehicle-road coupling model for closed-loop optimization, thereby achieving continuous improvement in the performance of the drive-by-wire chassis. In summary, this embodiment improves the adaptability, reliability, and development efficiency of the drive-by-wire chassis by constructing a technical system for intelligent optimization design of the drive-by-wire chassis, accelerates the maturity and perfection of the drive-by-wire chassis technology, and lays the foundation for industrial application.

[0115] In an optional embodiment,

[0116] The road roughness is estimated based on the intensity information of point cloud reflection as shown in the following formula:

[0117]

[0118] Where R represents the comprehensive roughness estimate, S represents the total number of scales, and w s is the weight coefficient, which indicates the importance of scale s in the overall roughness calculation, N is the total number of points in the point cloud, and K s represents the number of neighborhood points at scale s, Indicates point p under scale s i The normal vector of Indicates point p under scale s j The normal vector of is the normal vector and The inner product of represents the similarity between two normal vectors.

[0119] In this embodiment, a comprehensive roughness estimation value is obtained by calculating the similarity between normal vectors of point clouds at different scales, and the road roughness is converted from a qualitative description to a quantitative indicator, which provides accurate road attribute input for subsequent road parameterized modeling and vehicle-road coupling simulation, and improves the precision and reliability of the simulation. By introducing multi-scale analysis and calculating the similarity of point cloud normal vectors at different scales, the roughness characteristics of the road surface can be more comprehensively characterized, and the robustness and adaptability of the roughness estimation can be improved. By introducing weight coefficients, the importance of different scales can be adjusted according to actual needs, and the flexibility and adaptability of the roughness estimation can be improved. In summary, this embodiment provides a new idea for automatic evaluation of road roughness using on-board lidar point cloud data, and has good application prospects.

[0120] Figure 2 FIG. 1 is a schematic diagram of the structure of a chassis control-by-wire simulation system for an intelligent networked vehicle based on multi-domain modeling according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0121] The first unit is used to obtain structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, and at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamics equation is established, and the multi-flexible body dynamics equation is solved by a Newmark implicit and explicit algorithm to obtain vibration responses and stress-strain distributions under different working conditions;

[0122] The second unit is used to embed the multi-flexible body dynamics model into a preset drive-by-wire chassis integrated simulation model, perform multi-physics field finite element modeling through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, generate multiple finite element models, perform multi-physics field fusion at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, generate a multi-physics field model and integrate it into the drive-by-wire chassis integrated simulation model, build a multi-error source fusion model corresponding to multiple sensors, perform online error identification and active compensation through the Kalman filter algorithm, generate sensor data fusion results, feed back the sensor data fusion results, vibration responses and stress-strain distribution under different working conditions to the controller, perform closed-loop control of the drive-by-wire chassis under all working conditions in combination with the dynamic response characteristics of the drive-by-wire chassis, and generate a closed-loop integrated simulation model of the drive-by-wire chassis;

[0123] The third unit is used to collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parameterized model of the road, embed the three-dimensional parameterized model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full working condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

[0124] According to a third aspect of the embodiments of the present invention,

[0125] An electronic device is provided, comprising:

[0126] processor;

[0127] a memory for storing processor-executable instructions;

[0128] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0129] A fourth aspect of the embodiments of the present invention is:

[0130] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0131] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0132] Finally, it should be noted that the above 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 with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A chassis control-by-wire simulation method for intelligent connected vehicles based on multi-domain modeling, characterized in that: include: Acquire structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, introduce a coupling coordination equation at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model through the Lagrange multiplier method and establish a multi-flexible body dynamics equation, solve the multi-flexible body dynamics equation through the Newmark implicit explicit algorithm, and obtain vibration response and stress-strain distribution under different working conditions; The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, and multiple finite element models are generated. Multi-physics field fusion is performed at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, and a multi-physics field model is generated and integrated into the drive-by-wire chassis integrated simulation model. A multi-error source fusion model corresponding to multiple sensors is constructed, and online error identification and active compensation are performed through the Kalman filter algorithm to generate sensor data fusion results. The sensor data fusion results, vibration responses and stress-strain distributions under different working conditions are fed back to the controller, and closed-loop control of the drive-by-wire chassis under all working conditions is performed in combination with the dynamic response characteristics of the drive-by-wire chassis to generate a closed-loop integrated simulation model of the drive-by-wire chassis. Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parameterized model of the road, embed the three-dimensional parameterized model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model, and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full working condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

2. The method according to claim 1, characterized in that: The structural parameters and dynamic parameters of the intelligent networked vehicle chassis are obtained, a finite element mesh is constructed based on a parametric finite element method of material nonlinearity and collective nonlinearity, the finite element mesh is adaptively optimized, and a multi-flexible body dynamic model corresponding to the intelligent networked vehicle chassis is established, wherein the multi-flexible body dynamic model includes a flexible body dynamic sub-model and a rigid body dynamic sub-model, and at the coupling interface between the flexible body dynamic sub-model and the rigid body dynamic sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamic equation is established, and the multi-flexible body dynamic equation is solved by a Newmark implicit explicit algorithm, and the vibration response and stress-strain distribution under different working conditions are obtained, including: Reconstruct the surface profile information of different components in the intelligent networked vehicle chassis through three-dimensional scanning and structured light through reverse engineering methods to obtain a three-dimensional mesh model, and obtain the structural parameters corresponding to the intelligent networked vehicle chassis based on the three-dimensional mesh model, wherein the structural parameters include the size parameters, posture parameters and connection parameters corresponding to different components, and obtain the dynamic parameters corresponding to the intelligent networked vehicle chassis, wherein the dynamic parameters include mass, stiffness, damping and moment of inertia; Based on the structural parameters and the dynamic parameters, a finite element mesh corresponding to the chassis of the intelligent networked vehicle is constructed based on a parametric finite element method that takes into account material nonlinearity and collective nonlinearity; based on topological parameters corresponding to chassis configuration features, a logical relationship between chassis components and interfaces is determined and a consistent topological description is formed; based on the consistent topological description and the finite element mesh, corresponding gradient information of stress and strain is obtained through post-processing analysis; the gradient information is used as an error estimation indicator; the area and degree of local mesh refinement and local mesh coarsening are adaptively determined; a mesh refinement criterion and a mesh coarsening criterion are generated at the same time; and adaptive mesh optimization is performed based on the mesh refinement criterion and the mesh coarsening criterion; A multi-flexible body dynamics model corresponding to the intelligent connected vehicle chassis is established based on the optimized finite element mesh, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics model, the flexible body dynamics sub-model determines the large deformation and large rotation corresponding to the flexible body according to the absolute node coordinate method, constructs an initial flexible body dynamics equation in combination with non-conservative forces and corresponding constraints, and performs dimensionality reduction through the modal superposition method to obtain the flexible body dynamics equation, the rigid body dynamics model is modeled through the Newton-Euler equations, defines the position vector of any point of the rigid body and determines the kinetic energy of the rigid body, establishes the translation equation and the rotation equation of the rigid body in combination with the Euler angle parameterized rotation matrix, and determines the rigid body dynamics equation based on the translation equation and the rotation equation; Based on the rigid body dynamics equation and the flexible body dynamics equation, the coupled displacement coordination conditions and velocity coordination conditions are defined by the Lagrange multiplier method, the augmented Lagrange equation is constructed in combination with the Lagrange multiplier, the partial derivatives of the generalized coordinates are calculated and the multi-flexible body dynamics equation is constructed, the multi-flexible body dynamics equation is solved by the Newmark implicit and explicit algorithm combined with the mixed integration strategy, and the vibration response and stress-strain distribution of the connected vehicle chassis under different working conditions are obtained.

3. The method according to claim 2, characterized in that The multi-flexible body dynamics equation is solved by using the Newmark implicit and explicit algorithm combined with the mixed integration strategy as shown in the following formula: Among them, q n+1 represents the generalized coordinate vector of the next time step n+1, q n Represents the generalized coordinate vector of the current time step, △t n Represents the adaptive step size of the current time step, q′ n Represents the velocity vector of the current time step, q″ n represents the acceleration vector of the current time step, q″(n+1) represents the acceleration vector of the next time step, and β represents the acceleration control parameter.

4. The method according to claim 1, characterized in that The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, and multiple finite element models are generated. Multi-physics field fusion is performed at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, and a multi-physics field model is generated and integrated into the drive-by-wire chassis integrated simulation model. A multi-error source fusion model corresponding to multiple sensors is constructed, and online error identification and active compensation are performed through the Kalman filter algorithm to generate sensor data fusion results. The sensor data fusion results, vibration responses and stress-strain distributions under different working conditions are fed back to the controller, and closed-loop control of the drive-by-wire chassis under all working conditions is performed in combination with the dynamic response characteristics of the drive-by-wire chassis. The generation of a closed-loop integrated simulation model of the drive-by-wire chassis includes: The multi-flexible body dynamics model is embedded into a preset drive-by-wire chassis integrated simulation model, multi-physics field finite element modeling is performed through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, an electromagnetic field finite element model is constructed by combining the Maxwell equations and the geometric structure of the motor, the electromagnetic field equations are solved by a meshing method to obtain the magnetic field distribution data inside the motor, a fluid field finite element model is constructed according to the fluid flow and fluid control equations inside the hydraulic cylinder, and the fluid field equations are solved to obtain the velocity field and pressure field corresponding to the fluid, solid mechanics control equations are selected according to the solid structure part of the actuator and a solid field finite element model is constructed, and the solid field equations are solved in combination with the geometric shape and load conditions of the solid structure part to obtain the displacement field and stress field of the solid structure part; Based on the electromagnetic field finite element model, the fluid field finite element model and the solid field finite element model, a coupling interface between an actuator and a transmission mechanism is determined, coupling conditions are defined at the coupling interface, a mathematical constraint relationship of the coupling interface is established by a penalty function method, each finite element model is solved at each time step based on the mathematical constraint relationship, multi-physics field fusion is achieved by information exchange and iterative correction at the coupling interface, and the multi-physics field model is obtained and integrated into the integrated simulation model of the drive-by-wire chassis; A corresponding measurement model is constructed based on the preset sensors, the sensor error sources are determined based on the measurement model, and the multi-error source fusion model is constructed through the Kalman filter algorithm, the measurement data of each sensor is acquired in real time, and online error identification and active compensation are performed according to the multi-error source fusion model, the data after active compensation are used as the sensor data fusion result, combined with the vibration response and stress-strain distribution feedback and the dynamic response characteristics of the drive-by-wire chassis under the different working conditions, the closed-loop control of the drive-by-wire chassis under all working conditions is performed through a preset controller, and the closed-loop integrated simulation model of the drive-by-wire chassis is obtained.

5. The method according to claim 4, characterized in that The mathematical constraint relationship of the coupling interface is established by the penalty function method, and each finite element model is solved at each time step based on the mathematical constraint relationship as shown in the following formula: Among them, ε total is the overall energy function value, which represents the sum of the finite element model energy and penalty function, ε A is the energy of finite element model A, ε B represents the energy of the finite element model B, k i represents the penalty factor at level i, u A represents the displacement vector of the finite element model A on the coupling interface, u B represents the displacement vector of the finite element model B on the coupling interface, is an indicator function, indicating that when the displacement difference on the coupling interface is greater than the threshold ζ i 1 when ζ is set, otherwise 0. i Represents the displacement difference threshold of the i-th level.

6. The method according to claim 1, characterized in that Collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parametric model of the road, embed the three-dimensional parametric model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full-condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis, including: Collecting multi-source heterogeneous road map data corresponding to different cities, wherein the multi-source heterogeneous road map data includes high-precision maps, laser point clouds and street view images, obtaining the spatiotemporal references of different data sources based on inertial navigation technology and performing point cloud registration through control points, generating consistent point cloud data, and preprocessing the consistent point cloud data through a voxel filtering algorithm to obtain preprocessed multi-source heterogeneous road map data; For the pre-processed multi-source heterogeneous road map data, an end-to-end segmentation model is selected to construct a labeled data set for road semantic segmentation, wherein the labeled data set includes road images under different weather and lighting conditions, pixel-level annotation truth values ​​are generated by combining manual annotation and automatic generation, training is performed in combination with a cross entropy loss function and the end-to-end segmentation model is optimized by an optimizer, and a performance evaluation is performed using a pre-set independent test set. If the end-to-end segmentation model meets the pre-set performance standard, the end-to-end segmentation model is used to perform semantic segmentation on the processed multi-source heterogeneous road map data, identify urban roads, and extract road center lines and boundary lines as geometric features through a feature extraction algorithm, estimate road surface roughness and road surface material based on the intensity information of point cloud reflection and use them as road surface attribute parameters, extract environmental information corresponding to different roads based on the street view images, and extract environmental parameters through a feature extraction algorithm, and construct a three-dimensional parameterized road model based on the geometric features, road surface attribute parameters, and environmental parameters, and embed the three-dimensional parameterized road model into the closed-loop integrated simulation model of the drive-by-wire chassis to generate a vehicle-road coupling model; Based on the vehicle-road coupling model, the high-dimensional parameter space corresponding to the three-dimensional parameterized model of the road is explored through an intelligent sampling method, and the influence of different dimensional parameters on vehicle performance is evaluated in combination with a sensitivity analysis method and used as a sensitivity analysis result. Based on the sensitivity analysis result, key parameters and key parameter variation ranges are determined, and the sampling point distribution density is adjusted through an adaptive encryption strategy and simulation test cases are generated according to the sampling points. Based on the simulation test cases and a pre-set high-performance computing cluster, simulation test cases are executed in parallel, and combined with a task scheduling algorithm, simulation tests are performed with the goal of minimizing the test execution time, and the adaptability evaluation results corresponding to the drive-by-wire chassis are obtained; A full-operating-condition adaptability evaluation index system is established, including dynamic performance index, smoothness index, energy consumption index and stability index; the performance shortcomings in the adaptability evaluation results are identified according to the full-operating-condition adaptability evaluation index system and combined with the vibration response and stress-strain distribution under different operating conditions; the association rules between different performance indicators and control parameters are identified through an association rule mining algorithm; a knowledge graph and an optimization rule base are constructed based on the association rules and performance indicators; control parameter optimization suggestions corresponding to the current performance shortcomings are generated based on the optimization rule base and returned to the drive-by-wire chassis integrated simulation model; and the drive-by-wire chassis optimization design parameters and optimal control strategy are generated by repeated iterations.

7. The method according to claim 6, characterized in that The road roughness is estimated based on the intensity information of point cloud reflection as shown in the following formula: Where R represents the comprehensive roughness estimate, S represents the total number of scales, and w s is the weight coefficient, which indicates the importance of scale s in the overall roughness calculation, N is the total number of points in the point cloud, and K s represents the number of neighborhood points at scale s, Indicates point p under scale s i The normal vector of Indicates point p under scale s j The normal vector of is the normal vector and The inner product of represents the similarity between two normal vectors.

8. An intelligent connected vehicle chassis control-by-wire simulation system based on multi-domain modeling, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first unit is used to obtain structural parameters and dynamic parameters of the intelligent networked vehicle chassis, construct a finite element mesh based on a parametric finite element method of material nonlinearity and collective nonlinearity, adaptively optimize the finite element mesh and establish a multi-flexible body dynamics model corresponding to the intelligent networked vehicle chassis, wherein the multi-flexible body dynamics model includes a flexible body dynamics sub-model and a rigid body dynamics sub-model, and at the coupling interface between the flexible body dynamics sub-model and the rigid body dynamics sub-model, a coupling coordination equation is introduced by a Lagrange multiplier method and a multi-flexible body dynamics equation is established, and the multi-flexible body dynamics equation is solved by a Newmark implicit and explicit algorithm to obtain vibration responses and stress-strain distributions under different working conditions; The second unit is used to embed the multi-flexible body dynamics model into a preset drive-by-wire chassis integrated simulation model, perform multi-physics field finite element modeling through the drive-by-wire actuator subsystem in the drive-by-wire chassis integrated simulation model, generate multiple finite element models, perform multi-physics field fusion at the coupling interface between the actuator and the transmission mechanism through the fluid-solid coupling method, generate a multi-physics field model and integrate it into the drive-by-wire chassis integrated simulation model, build a multi-error source fusion model corresponding to multiple sensors, perform online error identification and active compensation through the Kalman filter algorithm, generate sensor data fusion results, feed back the sensor data fusion results, vibration responses and stress-strain distribution under different working conditions to the controller, perform closed-loop control of the drive-by-wire chassis under all working conditions in combination with the dynamic response characteristics of the drive-by-wire chassis, and generate a closed-loop integrated simulation model of the drive-by-wire chassis; The third unit is used to collect multi-source heterogeneous road map data corresponding to different cities, perform semantic segmentation and feature fusion on the multi-source heterogeneous road map data through deep learning methods, extract geometric parameters, road surface attribute parameters and environmental parameters corresponding to the road to construct a three-dimensional parameterized model of the road, embed the three-dimensional parameterized model of the road into the closed-loop integrated simulation model of the drive-by-wire chassis, generate a vehicle-road coupling model and generate simulation test cases through intelligent sampling and adaptive encryption methods, perform full working condition evaluation based on the simulation test cases, obtain adaptability evaluation results corresponding to the drive-by-wire chassis, identify performance shortcomings in the adaptability evaluation results and combine the vibration response and stress-strain distribution under different working conditions, mine the drive-by-wire strategy through machine learning methods and construct a knowledge graph and optimization rule base, generate control parameter optimization suggestions and return them to the drive-by-wire chassis integrated simulation model, and repeatedly iterate to generate optimized design parameters and optimal control strategies for the drive-by-wire chassis.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images