Wheel-ground relation modeling method and system for multi-wheel special vehicle in field environment

Abstract

The invention discloses a wheel-ground relation modeling method and system for a multi-wheel special vehicle in a field environment, and relates to the technical field of vehicle motion control. The method comprises a data acquisition step, a model construction step, a parameter identification step and a model verification step. According to the method, a virtual soft road surface data acquisition platform is established, and a wheel-ground contact model suitable for the soft road surface is established by aiming at the phenomenon that a multi-wheel special vehicle passes through the soft road surface for multiple times and taking accurate acquisition of lateral stress of each tire as a target; therefore, more reliable vehicle dynamic parameters are provided for establishment of a vehicle model and operation of a vehicle control system.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle motion control, and in particular to a wheel-ground relationship modeling method and system for a multi-wheeled special vehicle in a field environment. Background Art

[0002] Tire lateral force is crucial for vehicle lateral dynamics control, such as yaw stability control and trajectory tracking. For vehicles operating in the wild, the deformability of unstructured roads complicates the wheel-ground contact relationship. Special contact forces such as bulldozing resistance and compaction resistance generated by tire sinking cause the tire's mechanical properties to exhibit strong nonlinear characteristics. For multi-wheeled special-purpose vehicles, the interaction between the front wheels and the ground leaves ruts, altering parameters such as soil density and cohesion, directly affecting the rear wheels' ability to re-pass. This multiple-pass effect distorts the rear wheel lateral force characteristics, resulting in severe asymmetry. These factors collectively make it impossible to accurately capture tire lateral forces using traditional tire models for multi-wheeled special-purpose vehicles operating in the wild.

[0003] At present, the main method for fitting tire forces is to obtain the true value of tire forces during driving through experiments and compare it with the model calculated value. The main problems of this method are data collection and model correction and identification. 1) Data Collection: Field environments are expensive to build, require a large number of sensor facilities, and soil parameters are difficult to measure. Environmental uncertainty introduces a large amount of noise into the collection of true data, making it impossible to create an ideal experimental environment. Accurate experimental data is essential for building accurate models. 2) Model Modification and Identification: Currently, no research has developed a mechanistic tire model suitable for field conditions, let alone a model that considers the multipass effect. The structure of a semi-empirical tire model needs to be modified based on actual needs, and the parameters to be fitted determine the model's performance.

[0004] Therefore, it is an urgent problem for those skilled in the art to propose a wheel-ground relationship modeling method and system for multi-wheeled special vehicles in outdoor environments to solve the problems existing in the prior art. Summary of the Invention

[0005] In view of this, the present invention provides a wheel-ground relationship modeling method and system for multi-wheeled special vehicles in outdoor environments, builds a virtual soft road data acquisition platform, and aims to accurately obtain the lateral force of each tire in response to the phenomenon of multiple passages of multi-wheeled special vehicles on soft roads. A wheel-ground contact model and parameter set suitable for soft roads are established, thereby providing more reliable vehicle dynamics parameters for the establishment of vehicle models and the operation of vehicle control systems.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: A wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment comprises the following steps: S1 Data Acquisition Steps: Build a soft road simulation environment, establish a road driver in-the-loop data collection environment within the soft road simulation environment, and build a data post-processing platform based on the established soft road driver in-the-loop data collection environment; S2 Model construction steps: Adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle to obtain a modified tire lateral force model suitable for soft roads; S3 parameter identification step: Based on the multi-wheel special vehicle driving data on soft roads obtained in S1 and the tire lateral force model obtained in S2, the particle swarm optimization algorithm is used to fit the tire lateral force model parameters, and a mapping relationship between soil parameters and tire lateral force model parameters is established; S4 Model Verification Step: Conduct a joint simulation experiment based on the soft road simulation environment, the driver-in-the-loop data acquisition environment, and the data post-processing platform to determine the error between the calculated and measured tire lateral force values.

[0007] In the above method, optionally, the step S1 of acquiring data specifically includes the following: S101 Soft Road Simulation Environment Construction According to Bekker theory, a homogenized soft road surface simulation environment composed of clay was constructed; S102 Driver-in-the-loop environment establishment Based on the simulation environment built by S101, the steering wheel is connected to simulate the driver's real driving input, simulating the real driving behavior of multi-wheeled special vehicles on the constructed soft road surface; S103 Data post-processing platform construction Based on the soft road simulation environment built by S101 and the soft road driver in-loop data acquisition environment built by S102, data is transmitted to the external recording and post-processing via Ethernet.

[0008] In the above method, the optional S2 model construction steps specifically include the following: The original Burckhardt coupled tire model based on lateral force is as follows:

[0009] in, represents the lateral force of the tire, is the vertical force on the tire, represents the slip ratio, Indicates the tire slip angle, Represents the parameters to be fitted in the model; There is a nonlinear relationship between tire lateral force and tire slip angle under soft road conditions, and the tire force on the rear wheel shows asymmetry. The parameters of the Burckhardt original coupled tire model based on lateral force are adjusted according to the sign of the tire slip angle, without retaining the The simplified model expression retains the parameters related to the tire slip angle, and finally the tire lateral force model suitable for soft roads is obtained, as follows:

[0010] in, d represents the tire force hysteresis coefficient, represents the first-order derivative of the tire slip angle, b represents the lateral force bias, Indicates the tire slip angle.

[0011] In the above method, the specific contents of the S3 parameter identification step are optional and are as follows: S301 particle swarm optimization initialization: Tire force hysteresis coefficient and lateral force bias Manual adjustment is used to adjust the tire lateral force model parameters As the optimization particle of PSO, In addition, due to the influence of multiple pass effects and load transfer, the model parameters are fitted in sections according to the positive and negative tire slip angles. The total number of initial particles is set to N , set the maximum number of iterations to , initialize the initial position and velocity of each particle through the initialization function, and get ,in, For the The initial position of each particle; For the The initial velocity of each particle; S302 calculates the fitness function: In each iteration, the fitness function needs to be evaluated based on the current position of the particle. The fitness function is as follows:

[0012] in, Indicates the Particle No. The fitness function value of the round iteration, Indicates that due to The particle in The value of the wheel is based on the formula Generate a series of tire lateral force values, Indicates the real value collected based on the simulation environment built by S11; S303 updates particle position and velocity: In each round of iteration, the fitness function value of each particle is calculated to obtain the optimal historical position of the particle at the current iteration number. , the global optimal position of the particle swarm , according to the gradient change direction of the fitness function, calculate the latest speed and latest position of the particle position change at the current iteration:

[0013]

[0014] in, Indicates the The particle in wheel position; Indicates the The particle in wheel speed; represents the inertia factor; represents the learning factor; Indicates the number of seeds, A random number between S304 adaptive weight coefficient: The adaptive weight automatically adjusts the inertia weight factor based on the degree of convergence and fitness value. When the fitness value is less than the average fitness value, the inertia weight factor is as follows:

[0015] When the fitness value is greater than the average fitness value, the inertia weight factor is as follows:

[0016] in, represents the inertia weight factor, Represent the maximum and minimum values ​​of the inertia weight factor, respectively. Represents the current fitness value, Represent the average fitness value and the minimum fitness value respectively; S305 determines whether the termination condition is met: During the iteration process, loop according to S302-S304. Reached the maximum number of iterations When , the optimal particle position under the current working condition is obtained, that is, the optimal model parameters under the working condition , that is, the model parameter identification is completed; S306 parameter set establishment: The soil parameters change in properties after each wheel pass, and the changes are calculated as a function of slip as follows:

[0017] in, represents the density of the soil, represents the parameters to be fitted, represents the wheel slip rate, Indicates the number of passes; cohesion and shear modulus The properties of also change, and its calculation formula is as follows:

[0018]

[0019] in, Indicates that the wheels pass through the same soil The cohesion of the soil after Indicates that the wheels pass through the same soil Shear modulus of soil after treatment; According to the data collection environment established by S1, tire force data under different soil parameter combinations are collected, and S301-S305 are looped to identify the tire lateral force model established by S2 to obtain a parameter set of soil parameters and tire lateral force model parameters.

[0020] The above method, optional, S401 joint simulation experiment: Start the co-simulation and collect tire slip angle and tire force signals; S402 real-time calculation of lateral tire forces: Calculate the lateral force of each tire in real time based on the tire lateral force model obtained in S202 and the tire lateral force model parameters obtained in S3; S403 error calculation: Based on the calculated tire lateral force value obtained in S402, the error between the calculated tire lateral force value and the measured tire lateral force value is calculated using the normalized root mean square error calculation formula:

[0021] in, represents the true value of tire force collected through the simulation platform, represents the calculated value obtained by the proposed model, n Indicates the number of sampling points.

[0022] A wheel-land relationship modeling system for a multi-wheeled special vehicle in a field environment, used for implementing any of the above-mentioned wheel-land relationship modeling methods for a multi-wheeled special vehicle in a field environment, comprising a data acquisition module, a model construction module, a parameter identification module, and a model verification module connected in sequence; The data acquisition module builds a soft road simulation environment, establishes a road driver in-the-loop data collection environment in the soft road simulation environment, and builds a data post-processing platform based on the built soft road driver in-the-loop data collection environment; A model building module is used to adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle, thereby correcting the tire lateral force model suitable for soft roads; A parameter identification module is used to fit the tire lateral force model parameters using a particle swarm optimization algorithm based on the multi-wheel special vehicle driving data on soft roads obtained by the data acquisition module and the tire lateral force model obtained by the model construction module, and to establish a mapping relationship between soil parameters and tire lateral force model parameters; Model validation is used to conduct joint simulation experiments based on a soft road simulation environment, a driver-in-the-loop data acquisition environment, and a data post-processing platform to obtain the error between the calculated and measured tire lateral force.

[0023] As can be seen from the above technical solutions, compared with the prior art, the present invention provides a wheel-ground relationship modeling method and system for multi-wheeled special vehicles in outdoor environments, which has the following beneficial effects: The present invention builds a virtual soft road surface data acquisition platform. Aiming at the phenomenon of multiple passes of multi-wheeled special vehicles on soft roads, the present invention aims to accurately obtain the lateral force of each tire and establish a wheel-ground contact model and parameter set suitable for soft roads, thereby providing more reliable vehicle dynamics parameters for the establishment of vehicle models and the operation of vehicle control systems. By considering the influence of multiple pass effects on the lateral force characteristics of the front and rear wheels, the tire force output by the constructed model can effectively improve the accuracy of the vehicle dynamics model, so that the vehicle motion state can be predicted more accurately, providing high-quality input for the controller, and ultimately achieving better control effects and ensuring driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0025] Figure 1This is a flow chart of a wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment disclosed by the present invention; Figure 2 This is a schematic diagram of a wheel-ground relationship modeling system framework for a multi-wheeled special vehicle in a field environment disclosed in the present invention; Figure 3 This is a flow chart of a wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment disclosed in an embodiment of the present invention; Figure 4 Graphs illustrating the difference in front and rear tire force characteristics caused by the multi-pass effect disclosed in an embodiment of the present invention, where 4a represents the front wheel lateral force characteristic as a function of its sideslip angle, and 4b represents the rear wheel lateral force characteristic as a function of its sideslip angle; Figure 5 This is a real-time fitting effect diagram of the front and rear wheel lateral forces disclosed in an embodiment of the present invention, where 5a represents the fitting effect of the model for the front wheel lateral force, and 5b represents the fitting effect of the model for the rear wheel lateral force. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0027] In this application, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. The terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or apparatus comprising the element.

[0028] See also Figure 1 As shown, the present invention discloses a wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment, comprising the following steps: S1 Data Acquisition Steps: Build a soft road simulation environment, establish a road driver in-the-loop data collection environment within the soft road simulation environment, and build a data post-processing platform based on the established soft road driver in-the-loop data collection environment; S2 Model construction steps: Adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle to obtain a modified tire lateral force model suitable for soft roads; S3 parameter identification step: Based on the multi-wheel special vehicle driving data on soft roads obtained in S1 and the tire lateral force model obtained in S2, the particle swarm optimization algorithm is used to fit the tire lateral force model parameters, and a mapping relationship between soil parameters and tire lateral force model parameters is established; S4 Model Verification Step: Conduct a joint simulation experiment based on the soft road simulation environment, the driver-in-the-loop data acquisition environment, and the data post-processing platform to determine the error between the calculated and measured tire lateral force values.

[0029] Furthermore, data collection experiments conducted in the field usually consume a lot of resources and are difficult to repeat, which makes the creation of a virtual soft road data collection environment a necessary task. Specifically, the S1 data acquisition steps include the following: S101 Soft Road Simulation Environment Construction According to Bekker theory, a homogenized soft road surface simulation environment composed of clay was constructed; S102 Driver-in-the-loop environment establishment Based on the simulation environment built by S101, a steering wheel was connected to simulate the driver's real driving input (during the data collection process, the driver controlled the vehicle in the simulation environment using a Logitech G29 steering wheel), simulating the real driving behavior of multi-wheeled special vehicles on the constructed soft road surface; S103 Data post-processing platform construction Based on the soft road simulation environment built by S101 and the soft road driver in-loop data acquisition environment built by S102, data is transmitted to the external recording and post-processing via Ethernet.

[0030] Furthermore, semi-empirical formulas focus on the shape characteristics of the curve being fitted and directly derive mathematical expressions that accurately describe the system characteristics through curve fitting. The Burckhardt model has been shown to satisfy strong nonlinear relationships and is commonly used in tire modeling. Specifically, the S2 model construction steps include the following: The original Burckhardt coupled tire model based on lateral force is as follows:

[0031] in, represents the lateral force of the tire, is the vertical force on the tire, represents the slip ratio, Indicates the tire slip angle, Represents the parameters to be fitted in the model; Data collected by the S1 revealed a significant nonlinear relationship between tire lateral force and slip angle under soft road conditions, with the rear tire force exhibiting significant asymmetry. This made the constant-parameter Burckhardt model inadequate, requiring model parameter adjustments based on the sign of the slip angle. Based on the classic Burckhardt model, a modified Burckhardt model for multi-wheeled vehicles suitable for soft roads was developed.

[0032] There is a nonlinear relationship between tire lateral force and tire slip angle under soft road conditions, and the tire force on the rear wheel shows asymmetry. The parameters of the Burckhardt original coupled tire model based on lateral force are adjusted according to the sign of the tire slip angle, without retaining the The simplified model expression retains the parameters related to the tire side slip angle: Increase item, simulating tire force hysteresis.

[0033] Retention coefficient , which determines the magnitude of the lateral saturation force. The multiple-pass effect causes the front and rear wheels to encounter different terrain conditions, resulting in different lateral saturation forces, which in turn leads to significant differences in parameters between the front and rear wheels.

[0034] Retention coefficient , characterizes the lateral stiffness of the tire, determines the slope of the origin, is affected by the multiple-pass effect, and the rear wheel shows obvious asymmetry due to the destruction of terrain parameters, so a segmented form should be adopted.

[0035] Retain parameters , determines the saturation slope, especially when the axle load on the rear wheel decreases, the tire lateral force is very easy to saturate, and the influence of this parameter is extremely significant.

[0036] Add parameters , represents the value in static state, reflecting the lateral compaction resistance of the soil.

[0037] Finally, the tire lateral force model suitable for soft roads is obtained as follows:

[0038] in, d represents the tire force hysteresis coefficient, represents the first-order derivative of the tire slip angle, b represents the lateral force bias, Indicates the tire slip angle.

[0039] Furthermore, the particle swarm algorithm randomly distributes the initial particle swarm in space, simulating the foraging behavior of the population and finding the optimal solution in the entire space based on the group's and its own experience. Specifically, the S3 parameter identification step is as follows: S301 particle swarm optimization initialization: Tire force hysteresis coefficient and lateral force bias Manual adjustment is used to adjust the tire lateral force model parameters As the optimization particle of PSO, In addition, due to the influence of multiple pass effects and load transfer, the model parameters are fitted in sections according to the positive and negative tire slip angles. The total number of initial particles is set to N, and the maximum number of iterations is set to , initialize the initial position and velocity of each particle through the initialization function, and get ,in, For the The initial position of each particle; For the The initial velocity of each particle; S302 calculates the fitness function: In each iteration, the fitness function needs to be evaluated based on the current position of the particle. The fitness function is as follows:

[0040] in, Indicates the Particle No. The fitness function value of the round iteration, Indicates that due to The particle in The value of the wheel is based on the formula Generate a series of tire lateral force values, Indicates the real value collected based on the simulation environment built by S11; S303 updates particle position and velocity: In each round of iteration, the fitness function value of each particle is calculated to obtain the optimal historical position of the particle at the current iteration number. , the global optimal position of the particle swarm , according to the gradient change direction of the fitness function, calculate the latest speed and latest position of the particle position change at the current iteration:

[0041]

[0042] in, Indicates the The particle in wheel position; Indicates the The particle in wheel speed; represents the inertia factor; represents the learning factor; Indicates the number of seeds, A random number between S304 adaptive weight coefficient: The inertia factor has a great influence on the ability of the particle swarm algorithm to search for the optimal solution. When the inertia factor is large, the global search ability of the particle swarm algorithm is strong and the search speed is fast, but the search result accuracy is low. When the inertia factor is small, the search result accuracy of the particle swarm algorithm is high, but the search speed is slow and it is more likely to fall into the local optimal solution.

[0043] The adaptive weight automatically adjusts the inertia weight factor based on the degree of convergence and fitness value. When the fitness value is less than the average fitness value, the inertia weight factor is as follows:

[0044] When the fitness value is greater than the average fitness value, the inertia weight factor is as follows:

[0045] in, represents the inertia weight factor, Represent the maximum and minimum values ​​of the inertia weight factor, respectively. Represents the current fitness value, Represent the average fitness value and the minimum fitness value respectively; S305 determines whether the termination condition is met: During the iteration process, loop according to S302-S304. Reached the maximum number of iterations When , the optimal particle position under the current working condition is obtained, that is, the optimal model parameters under the working condition , that is, the model parameter identification is completed; S306 parameter set establishment: The soil parameters change in properties after each wheel pass, and the changes are calculated as a function of slip as follows:

[0046] in, represents the density of the soil, represents the parameters to be fitted, represents the wheel slip rate, Indicates the number of passes; cohesion and shear modulus The properties of also change, and its calculation formula is as follows:

[0047]

[0048] in, Indicates that the wheels pass through the same soil The cohesion of the soil after Indicates that the wheels pass through the same soil Shear modulus of soil after treatment; Repeated compaction of the soil causes changes in internal parameters, which in turn causes changes in the wheel-ground interaction relationship between the front and rear wheels of multi-wheeled special vehicles.

[0049] For the mechanism-based tire model, soil parameters will directly affect the model calculation value. For the semi-empirical tire model, changes in soil parameters will cause changes in model parameters, which in turn affect the model calculation value.

[0050] According to the data collection environment established by S1, tire force data under different soil parameter combinations are collected, and S301-S305 are looped to identify the tire lateral force model established by S2 to obtain a parameter set of soil parameters and tire lateral force model parameters.

[0051] Further, S401 joint simulation experiment: Start the co-simulation and collect tire slip angle and tire force signals; S402 real-time calculation of lateral tire forces: Calculate the lateral force of each tire in real time based on the tire lateral force model obtained in S202 and the tire lateral force model parameters obtained in S3; S403 error calculation: Based on the calculated tire lateral force value obtained in S402, the error between the calculated tire lateral force value and the measured tire lateral force value is calculated using the normalized root mean square error calculation formula:

[0052] in, represents the true value of tire force collected through the simulation platform, represents the calculated value obtained by the proposed model, n Indicates the number of sampling points.

[0053] and Figure 1 Corresponding to the method shown, the present invention also discloses a wheel-ground relationship modeling system for multi-wheeled special vehicles in outdoor environments, which is used Figure 1The implementation of a wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment shown in the figure includes a data acquisition module, a model construction module, a parameter identification module, and a model verification module connected in sequence; The data acquisition module is used to build a soft road simulation environment, establish a road driver in the loop data collection environment in the soft road simulation environment, and build a data post-processing platform based on the built soft road driver in the loop data collection environment; A model building module is used to adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle, thereby correcting the tire lateral force model suitable for soft roads; A parameter identification module is used to fit the tire lateral force model parameters using a particle swarm optimization algorithm based on the multi-wheel special vehicle driving data on soft roads obtained by the data acquisition module and the tire lateral force model obtained by the model construction module, and to establish a mapping relationship between soil parameters and tire lateral force model parameters; The model validation module conducts joint simulation experiments using a soft road simulation environment, a driver-in-the-loop data acquisition environment, and a data post-processing platform to determine the error between the calculated and measured tire lateral forces. Based on the validation results, a tire model suitable for multi-wheeled vehicles on soft roads is developed. Experimental results demonstrate that the proposed model accurately fits the lateral forces of the vehicle's front and rear wheels.

[0054] In a specific embodiment, the present invention builds a virtual soft road data acquisition platform. Targeting the phenomenon of multiple special vehicles passing through soft roads, the present invention aims to accurately obtain the lateral forces on each tire and establish a wheel-ground contact model suitable for soft roads, thereby providing more reliable vehicle dynamics parameters for the establishment of vehicle models and the operation of vehicle control systems. Figure 3 As shown, the steps involved in the method of this embodiment are as follows: Step 1: Data Collection 1.1 Chrono soft road simulation environment construction Compared to other dynamics simulation tools like Adams and CarSim, Chrono's Soil Contact Model (SCM) stands out for its efficient data processing, providing real-time or near-real-time simulation capabilities and accurately simulating soil deformation and changes in soil parameters resulting from wheel-ground interaction. These features make the SCM an ideal platform for studying vehicle dynamics on deformable terrain.

[0055] Based on Bekker's theory, a homogenized soft road surface made of clay was constructed using Chrono. Its main parameters are shown in Table 1: Table 1

[0056] 1.2 Logitech G29 Driver-in-the-Loop Environment Establishment

[0057] Based on the simulation environment built in step 1.1, connect the Logitech G29 steering wheel through the Logitech official driver to simulate the driver's real driving input and simulate the real driving behavior of multi-wheeled special vehicles on the constructed soft road surface.

[0058] 1.3 Simulink data post-processing platform construction

[0059] Based on the soft road driver-in-the-loop data acquisition environment built in steps 1.1 and 1.2, the data is transmitted to Simulink via Ethernet for recording and post-processing.

[0060] Step 2: Model building

[0061] Semi-empirical formulas focus on the shape characteristics of the curve being fitted and directly derive mathematical expressions that accurately describe the system characteristics through curve fitting. The Burckhardt model has been shown to satisfy strong nonlinear relationships and is commonly used in tire modeling. The original Burckhardt coupled tire equation for lateral force is given as follows:

[0062] in, is the lateral force on the tire, is the vertical force on the tire, is the slip ratio, is the tire slip angle, are the parameters to be fitted in the model.

[0063] The model parameters are adjusted according to the sign of the tire slip angle. Based on the classic Burckhardt model, a modified Burckhardt model for multi-wheeled vehicles on soft roads is obtained.

[0064] Not retained The simplified model expression is only related to the tire slip angle.

[0065] Increase item, simulating tire force hysteresis.

[0066] Retention coefficient , which determines the magnitude of the lateral saturation force. The multiple-pass effect causes the front and rear wheels to encounter different terrain conditions, resulting in different lateral saturation forces, which in turn leads to significant differences in parameters between the front and rear wheels.

[0067] Retention coefficient , characterizes the lateral stiffness of the tire, determines the slope of the origin, is affected by the multiple-pass effect, and the rear wheel shows obvious asymmetry due to the destruction of terrain parameters, so a segmented form should be adopted.

[0068] Retain parameters , determines the saturation slope, especially when the axle load on the rear wheel decreases, the tire lateral force is very easy to saturate, and the influence of this parameter is extremely significant.

[0069] Add parameters , represents the value in static state, reflecting the lateral compaction resistance of the soil.

[0070] The final tire lateral force model is as follows:

[0071] in, d represents the tire force hysteresis coefficient, represents the first-order derivative of the tire slip angle, b represents the lateral force bias, Indicates the tire slip angle.

[0072] Step 3: Parameter identification

[0073] Based on the multi-wheel special vehicle driving data on soft roads obtained in step 1 and the tire lateral force model obtained in step 2, the particle swarm optimization algorithm is used to fit the model parameters and establish a mapping relationship between soil parameters and model parameters.

[0074] The particle swarm algorithm randomly distributes the initial particle swarm in the space and searches for the optimal solution in the entire space by simulating the foraging behavior of the population based on the group and its own experience.

[0075] 3.1 Particle Swarm Optimization Initialization

[0076] Tire force hysteresis coefficient and lateral force bias The adjustment is relatively easy and can be adjusted manually, so the tire model parameters are As the optimization particle of PSO, In addition, due to the influence of multiple-pass effect and load transfer, the model parameters should be fitted in sections according to the positive and negative tire slip angles.

[0077] The total number of initial particles is set to N, and the maximum number of iterations is set to , initialize the initial position and velocity of each particle through the initialization function, and get . For the The initial position of each particle; For the The initial velocity of the particle.

[0078] 3.2 Calculating the fitness function

[0079] In each iteration, the fitness function needs to be evaluated based on the current position of the particle. The fitness function of the present invention is selected as:

[0080] in For the Particle No. The fitness function value of the round iteration, It is because of the The particle in The value of the wheel is based on the formula Generate a series of tire lateral force values, This is the true value collected from the simulation environment built in step 1.

[0081] 3.3 Update particle position and velocity

[0082] In each round of iteration, by calculating the fitness function value of each particle, the optimal historical position of the particle at the current iteration number can be obtained. , the global optimal position of the particle swarm , according to the gradient change direction of the fitness function, calculate the latest speed and latest position of the particle position change at the current iteration:

[0083]

[0084] in, For the The particle in wheel position; For the The particle in wheel speed; is the inertia factor; is the learning factor; is the number of seeds, A random number between .

[0085] 3.4 Adaptive Weight Coefficient

[0086] The inertia factor has a great influence on the ability of the particle swarm algorithm to search for the optimal solution. When the inertia factor is large, the global search ability of the particle swarm algorithm is strong and the search speed is fast, but the search result accuracy is low. When the inertia factor is small, the search result accuracy of the particle swarm algorithm is high, but the search speed is slow and it is more likely to fall into the local optimal solution.

[0087] Adaptive weight can automatically adjust the inertia weight factor according to the degree of convergence and fitness value. When the fitness value is less than the average fitness value:

[0088] When the fitness value is greater than the average fitness value:

[0089] in, is the inertia weight factor, are the maximum and minimum values ​​of the inertia weight factor, is the current fitness value, are the average fitness value and the minimum fitness value respectively.

[0090] 3.5 Determine whether the termination condition is met

[0091] During the iteration process, follow steps 3.2-3.4 to loop. Reached the maximum number of iterations When , the optimal particle position under the current working condition is obtained, that is, the optimal model parameters under the working condition , that is, the model parameter identification is completed.

[0092] 3.6 Parameter Set Establishment

[0093] The soil parameters change their properties after each pass and this change is a function of the slip

[0094] in, represents the density of the soil, is the parameter to be fitted, is the wheel slip ratio, is the number of passes.

[0095] Similarly, cohesion and shear modulus There are also similar features:

[0096]

[0097] Repeated compaction of the soil causes changes in internal parameters, which in turn causes changes in the wheel-ground interaction relationship between the front and rear wheels of multi-wheeled special vehicles.

[0098] For the mechanism-based tire model, soil parameters will directly affect the model calculation value. For the semi-empirical tire model, changes in soil parameters will cause changes in model parameters, which in turn affect the model calculation value.

[0099] According to the data collection environment established in step 1, tire force data under different soil parameter combinations are collected. Steps 3.1-3.5 are repeated to identify the model established in step 2 to obtain a parameter set of soil parameters and model parameters.

[0100] Step 3: Model Validation

[0101] According to step 3, the lateral force models of the front and rear tires are obtained. According to step 1, the experimental conditions such as soft road surface, driver-in-the-loop, and data post-processing platform are set up to conduct a Chrono-Simulink joint simulation experiment.

[0102] 4.1 Joint Simulation Experiment

[0103] Start the Chrono-Simulink co-simulation. In this example, the vehicle is running on a soft road at a speed of During the serpentine maneuver, when the road conditions remain unchanged, the main signals collected are tire slip angle and tire force.

[0104] 4.2 Real-time calculation of lateral tire forces

[0105] According to the tire model expression obtained in step 2.2 and the model parameters obtained in steps 3.1-3.5, Simulink is used to calculate the lateral force of each tire in real time.

[0106] 4.3 Error calculation

[0107] Based on the tire force calculation value obtained in step 4.2, refer to the normalized root mean square error calculation formula to calculate the percentage error between the calculated value and the tire force measurement value: .

[0108] The Burckhardt model for multi-wheeled vehicles on soft roads constructed in step 2 does not have an integral form compared to the traditional Bekker model. Its structure is simple and can be used for controller design. This model also considers the influence of the interaction between the front wheel and the ground on the force characteristics of the rear tire, which can more accurately fit the tire force and characterize the vehicle dynamics. The completed experimental results are as follows: Figure 4 and Figure 5 As shown, from Figure 4 As can be seen from Figures 4a and 4b, the lateral force characteristics of the front and rear wheels are significantly different due to the multiple-pass effect. The hysteresis of the rear wheel is more obvious and the saturation speed is faster. Figure 5 5a and 5b, it can be seen that the proposed model can calculate the front and rear wheel lateral forces in real time with high accuracy.

[0109] As for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For relevant parts, please refer to the partial description of the method embodiment. The system and system embodiment described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without expending creative work.

[0110] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment, characterized in that: The following steps are involved: S1 Data Acquisition Steps: Build a soft road simulation environment, establish a road driver in-the-loop data collection environment within the soft road simulation environment, and build a data post-processing platform based on the established soft road driver in-the-loop data collection environment; S2 Model construction steps: Adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle to obtain a modified tire lateral force model suitable for soft roads; S3 parameter identification step: Based on the multi-wheel special vehicle driving data on soft roads obtained in S1 and the tire lateral force model obtained in S2, the particle swarm optimization algorithm is used to fit the tire lateral force model parameters, and a mapping relationship between soil parameters and tire lateral force model parameters is established; S4 Model Verification Step: Conduct a joint simulation experiment based on the soft road simulation environment, the driver-in-the-loop data acquisition environment, and the data post-processing platform to determine the error between the calculated and measured tire lateral force values.

2. The wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment according to claim 1, characterized in that: The S1 data acquisition steps specifically include the following: S101 Soft Road Simulation Environment Construction According to Bekker theory, a homogenized soft road surface simulation environment composed of clay was constructed; S102 Driver-in-the-loop environment establishment Based on the simulation environment built by S101, the steering wheel is connected to simulate the driver's real driving input, simulating the real driving behavior of multi-wheeled special vehicles on the constructed soft road surface; S103 Data post-processing platform construction Based on the soft road simulation environment built by S101 and the soft road driver in-loop data acquisition environment built by S102, data is transmitted to the external recording and post-processing via Ethernet.

3. The wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment according to claim 2, characterized in that: The S2 model construction steps specifically include the following: The original Burckhardt coupled tire model based on lateral force is as follows: in, represents the lateral force of the tire, is the vertical force on the tire, represents the slip ratio, Indicates the tire slip angle, Represents the parameters to be fitted in the model; There is a nonlinear relationship between tire lateral force and tire slip angle under soft road conditions, and the tire force on the rear wheel shows asymmetry. The parameters of the Burckhardt original coupled tire model based on lateral force are adjusted according to the sign of the tire slip angle, without retaining the The simplified model expression retains the parameters related to the tire slip angle, and finally the tire lateral force model suitable for soft roads is obtained, as follows: in, d represents the tire force hysteresis coefficient, represents the first-order derivative of the tire slip angle, b represents the lateral force bias, Indicates the tire slip angle.

4. The wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment according to claim 3, characterized in that: The specific contents of the S3 parameter identification steps are as follows: S301 particle swarm optimization initialization: Tire force hysteresis coefficient and lateral force bias Manual adjustment is used to adjust the tire lateral force model parameters As the optimization particle of PSO, In addition, due to the influence of multiple pass effects and load transfer, the model parameters are fitted in sections according to the positive and negative tire slip angles. The total number of initial particles is set to N, and the maximum number of iterations is set to , initialize the initial position and velocity of each particle through the initialization function, and get ,in, For the The initial position of each particle; For the The initial velocity of each particle; S302 calculates the fitness function: In each iteration, the fitness function needs to be evaluated based on the current position of the particle. The fitness function is as follows: in, Indicates the Particle No. The fitness function value of the round iteration, Indicates that due to The particle in The value of the wheel is based on the formula Generate a series of tire lateral force values, Indicates the real value collected based on the simulation environment built by S11; S303 updates particle position and velocity: In each round of iteration, the fitness function value of each particle is calculated to obtain the optimal historical position of the particle at the current iteration number. , the global optimal position of the particle swarm , according to the gradient change direction of the fitness function, calculate the latest speed and latest position of the particle position change at the current iteration: in, Indicates the The particle in wheel position; Indicates the The particle in wheel speed; represents the inertia factor; represents the learning factor; Indicates the number of seeds, A random number between S304 adaptive weight coefficient: The adaptive weight automatically adjusts the inertia weight factor based on the degree of convergence and fitness value. When the fitness value is less than the average fitness value, the inertia weight factor is as follows: When the fitness value is greater than the average fitness value, the inertia weight factor is as follows: in, represents the inertia weight factor, Represent the maximum and minimum values ​​of the inertia weight factor, respectively. Represents the current fitness value, Represent the average fitness value and the minimum fitness value respectively; S305 determines whether the termination condition is met: During the iteration process, loop according to S302-S304. Reached the maximum number of iterations When , the optimal particle position under the current working condition is obtained, that is, the optimal model parameters under the working condition , that is, the model parameter identification is completed; S306 parameter set establishment: The soil parameters change in properties after each wheel pass, and the changes are calculated as a function of slip as follows: in, represents the density of the soil, represents the parameters to be fitted, represents the wheel slip rate, Indicates the number of passes; cohesion and shear modulus The properties of also change, and its calculation formula is as follows: in, Indicates that the wheels pass through the same soil The cohesion of the soil after Indicates that the wheels pass through the same soil Shear modulus of soil after treatment; According to the data collection environment established by S1, tire force data under different soil parameter combinations are collected, and S301-S305 are looped to identify the tire lateral force model established by S2 to obtain a parameter set of soil parameters and tire lateral force model parameters.

5. The wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment according to claim 4, characterized in that: S401 joint simulation experiment: Start the co-simulation and collect tire slip angle and tire force signals; S402 real-time calculation of lateral tire forces: Calculate the lateral force of each tire in real time based on the tire lateral force model obtained in S202 and the tire lateral force model parameters obtained in S3; S403 error calculation: Based on the calculated tire lateral force value obtained in S402, the error between the calculated tire lateral force value and the measured tire lateral force value is calculated using the normalized root mean square error calculation formula: in, represents the true value of tire force collected through the simulation platform, represents the calculated value obtained by the proposed model, n Indicates the number of sampling points.

6. A wheel-ground relationship modeling system for multi-wheeled special vehicles in outdoor environments, characterized in that: A method for implementing a wheel-ground relationship modeling method for a multi-wheeled special vehicle in a field environment as described in any one of claims 1 to 5, comprising a data acquisition module, a model construction module, a parameter identification module, and a model verification module connected in sequence; The data acquisition module builds a soft road simulation environment, establishes a road driver in-the-loop data collection environment in the soft road simulation environment, and builds a data post-processing platform based on the built soft road driver in-the-loop data collection environment; A model building module is used to adjust the parameters of the Burckhardt original coupled tire model based on lateral force according to the sign of the tire slip angle, thereby correcting the tire lateral force model suitable for soft roads; A parameter identification module is used to fit the tire lateral force model parameters using a particle swarm optimization algorithm based on the multi-wheel special vehicle driving data on soft roads obtained by the data acquisition module and the tire lateral force model obtained by the model construction module, and to establish a mapping relationship between soil parameters and tire lateral force model parameters; The model verification module is used to conduct joint simulation experiments based on the soft road simulation environment, the driver-in-the-loop data acquisition environment, and the data post-processing platform to obtain the error between the calculated tire lateral force and the measured tire lateral force.

Images