Modelling Method of Unmanned Vehicle Dynamics Based on Neural Network

doi:10.16182/j.issn1004731x.joss.24-0935

Abstract

Abstract:

To address the challenges of high data acquisition costs of test data on dynamic characteristics between tires and soft terrain and low speed of numerical calculation for unmanned vehicles in complex terrestrial environments, a modeling method of unmanned vehicle dynamics based on a neural network was proposed. Tire-terrain contact dynamics models were built by using discrete element method (DEM) simulations for tire-terrain contact and experimental data, thereby creating a dataset of tire contact forces for various tire materials in terrestrial environments. The neural network was applied to regressively learn the dataset, and a nonlinear neural network tire model was constructed. Subsequently, a three-degree-of-freedom dynamics model of unmanned vehicles was constructed, and a physically meaningful and bounded dataset was generated. The results demonstrate that the model can achieve high-precision and efficient simulation of unmanned vehicle dynamics, fulfill the requirements for large-scale, high-precision, and fast calculation in simulated environments, and achieve fine-grained trajectory tracking control of unmanned vehicles.

Key words: neural network, deep learning, automotive dynamics modeling, discrete element modeling of contact, tire friction modeling

CLC Number:

TP18

Wang Jun, Liu Min, Zhang Xiaochuan, Ding Yishan, Feng Juhui, Zhuang Ye. Modelling Method of Unmanned Vehicle Dynamics Based on Neural Network[J]. Journal of System Simulation, 2026, 38(4): 932-947.

Figures/Tables 20

Fig. 1

Fig. 2

Table 1

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Fig. 14

Fig. 15

Fig. 16

Table 2

Fig. 17

Fig. 18

References 21

[1]	Pan Yongjun, Nie Xiaobo, Li Zhixiong, et al. Data-driven Vehicle Modeling of Longitudinal Dynamics Based on a Multibody Model and Deep Neural Networks[J]. Measurement, 2021, 180: 109541.
[2]	Liu Siyuan, Liu Yancheng, Liang Xiaoling, et al. Uncertainty Observation-based Adaptive Succinct Fuzzy-neuro Dynamic Surface Control for Trajectory Tracking of Fully Actuated Underwater Vehicle System with Input Saturation[J]. Nonlinear Dynamics, 2019, 98(3): 1683-1699.
[3]	Funahashi Ken-ichi, Nakamura Yuichi. Approximation of Dynamical Systems by Continuous Time Recurrent Neural Networks[J]. Neural Networks, 1993, 6(6): 801-806.
[4]	Devineau Guillaume, Polack Philip, Altché Florent, et al. Coupled Longitudinal and Lateral Control of a Vehicle Using Deep Learning[C]//2018 21st International Conference on Intelligent Transportation Systems (ITSC). Piscataway: IEEE, 2018: 642-649.
[5]	Garimella G, Funke J, Wang Chuang, et al. Neural Network Modeling for Steering Control of an Autonomous Vehicle[C]//2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Piscataway: IEEE, 2017: 2609-2615.
[6]	Osman O A, Rakha H. Application of Deep Learning for Characterization of Drivers' Engagement in Secondary Tasks in In-Vehicle Systems[J]. Transportation Research Record, 2020, 2674(8): 429-440.
[7]	Spielberg N A, Brown M, Kapania N R, et al. Neural Network Vehicle Models for High-performance Automated Driving[J]. Science Robotics, 2019, 4(28): eaaw1975.
[8]	Spielberg N A, Brown M, Gerdes J C. Neural Network Model Predictive Motion Control Applied to Automated Driving with Unknown Friction[J]. IEEE Transactions on Control Systems Technology, 2022, 30(5): 1934-1945.
[9]	Rokonuzzaman Mohammad, Mohajer Navid, Nahavandi Saeid, et al. Model Predictive Control with Learned Vehicle Dynamics for Autonomous Vehicle Path Tracking[J]. IEEE Access, 2021, 9: 128233-128249.
[10]	庄晔. 轮胎动摩擦特性研究及其对车辆操纵稳定性的影响[D]. 长春: 吉林大学, 2004.
	Zhuang Ye. Study on Dynamic Friction Property of Tire and Its Effect on Vehicle Handling[D]. Changchun: Jilin University, 2004.
[11]	李翔宇, 么鸣涛, 李宝玉, 等. 基于MATLAB的汽车操纵稳定性仿真分析[J]. 汽车实用技术, 2023, 48(24): 95-100.
	Li Xiangyu, Yao Mingtao, Li Baoyu, et al. Simulation Analysis of Vehicle Handling and Stability Based on MATLAB[J]. Automobile Applied Technology, 2023, 48(24): 95-100.
[12]	LeCun Y, Bengio Y, Hinton G. Deep Learning[J]. Nature, 2015, 521(7553): 436-444.
[13]	柴慧敏, 张勇, 李欣粤, 等. 基于深度学习的空中目标威胁评估方法[J]. 系统仿真学报, 2022, 34(7): 1459-1467.
	Chai Huimin, Zhang Yong, Li Xinyue, et al. Aerial Target Threat Assessment Method Based on Deep Learning[J]. Journal of System Simulation, 2022, 34(7): 1459-1467.
[14]	孙辉辉, 胡春鹤, 张军国. 移动机器人运动规划中的深度强化学习方法[J]. 控制与决策, 2021, 36(6): 1281-1292.
	Sun Huihui, Hu Chunhe, Zhang Junguo. Deep Reinforcement Learning for Motion Planning of Mobile Robots[J]. Control and Decision, 2021, 36(6): 1281-1292.
[15]	杨路辉, 刘光杰, 翟江涛, 等. 一种改进的卷积神经网络恶意域名检测算法[J]. 西安电子科技大学学报, 2020, 47(1): 37-43.
	Yang Luhui, Liu Guangjie, Zhai Jiangtao, et al. Improved Algorithm for Detection of the Malicious Domain Name Based on the Convolutional Neural Network[J]. Journal of Xidian University, 2020, 47(1): 37-43.
[16]	胡琦瑶, 杨皓文, 王佳欣, 等. 基于弱监督深度学习的图像检索技术研究[J]. 西北大学学报(自然科学版), 2020, 50(5): 793-801.
	Hu Qiyao, Yang Haowen, Wang Jiaxin, et al. Research on Image Retrieval Based on Weakly-supervised Deep Learning[J]. Journal of Northwest University(Natural Science Edition), 2020, 50(5): 793-801.
[17]	He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep Residual Learning for Image Recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE. 2016: 770-778.
[18]	仝卫国, 曾世超, 张立峰, 等. 基于深度残差神经网络的电阻层析成像及流型辨识方法[J]. 系统仿真学报, 2022, 34(9): 2028-2036.
	Tong Weiguo, Zeng Shichao, Zhang Lifeng, et al. Electrical Resistance Tomography and Flow Pattern Identification Method Based on Deep Residual Neural Network[J]. Journal of System Simulation, 2022, 34(9): 2028-2036.
[19]	胥雪炎, 李补喜. 不同被解释变量选择对决定系数R²的影响研究[J]. 太原科技大学学报, 2007, 28(5): 363-365.
	Xu Xueyan, Li Buxi. Research on the Effect of Selection of Dependent Variables on R² Statistic[J]. Journal of Taiyuan University of Science and Technology, 2007, 28(5): 363-365.
[20]	柴同, 袁逸萍, 马军岩, 等. 基于K-CNN和N-GRU的风电机组发电机状态预测[J]. 机械强度, 2023, 45(5): 1043-1049.
	Chai Tong, Yuan Yiping, Ma Junyan, et al. State Prediction of Wind Turbine Generator Based On K-CNN and N-GRU[J]. Journal of Mechanical Strength, 2023, 45(5): 1043-1049.
[21]	Wool H E D. The Relation Between Measures of Correlation in the Universe of Sample Permutations[J]. Biometrika, 1944, 33(2): 129-135.

参数	沙土颗粒	粘土颗粒
形状	球形	球形
半径/mm	2	2
密度/(kg/m³)	2 100	1 550
剪切模量/MPa	83	16.5
泊松比	0.25	0.25
颗粒-颗粒恢复系数	0.48	0.58
颗粒-颗粒静摩擦系数	0.51	1.21
颗粒-颗粒动摩擦系数	0.71	0.13
颗粒-几何体恢复系数	0.5	0.48
颗粒-几何体静摩擦系数	0.5	0.55
颗粒-几何体动摩擦系数	0.3	0.37
含水量/%	15	55
变形模量	0.7	0.4
内聚模量/(kN/m ⁿ⁺¹)	5.27	16.03
摩擦模量/(kN/m ⁿ⁺¹)	1 515.04	126.53
粘聚力/kPa	1.72	2.07
内摩擦角/(°)	34	10
厚度/mm	300	300

输出状态量	残差神经网络	前馈神经网络	循环神经网络
横摆角速度	0.999 998	0.998 507	0.841 569
纵向速度	0.999 978	0.512 992	0.187 719
侧向速度	0.999 991	0.998 709	0.706 560
纵向加速度	0.999 991	0.996 608	0.177 136
侧向加速度	0.999 998	0.999 840	0.925 140