【作者】 刘少杰；

【导师】 靳立强；

【作者基本信息】 吉林大学 ， 机械（专业学位）， 2025， 硕士

【摘要】 随着车辆“无人化”进程不断加速,针对传统车辆构型的轨迹跟踪控制性能已经被充分利用和开发。自动轮技术突破车辆传统布局,赋予车辆新的行驶模式,拓展了车辆的极限性能。然而,基于传统车辆构型的轨迹跟踪控制技术难以发挥自动轮智行底盘车辆的性能优势,并且难以跟踪极限工况下的行驶轨迹。本文依托于国家重点研发项目(2021YFB2500703)“多系统高效集成轮毂电机行动模块与整车转矩矢量分配技术”和广西科技重大专项(桂科AA24206032)“分布式四驱驱动智能线控底盘控制技术”,以自动轮智行底盘车辆作为研究对象,进行汽车行驶期望轨迹跟踪的非线性稳定控制研究。主要研究工作如下:(1)车辆动力学建模与行驶期望轨迹生成。研究首先构建四轮独立驱动/制动和四轮独立转向的自动轮智行底盘车辆动力学模型,并考虑空气阻力的影响。采用Dugoff半经验轮胎模型,描述轮胎的非线性特性,并构建车轮的旋转动力学模型,以及计算一些关键参数。提出一种基于时空遗留风险模型的轨迹规划方法。在Frenet坐标系下,构建三次样条网络化道路,针对道路、主车道和障碍物建立时空遗留风险模型,提高对动态环境风险的表述能力。采用动态规划求解期望行驶轨迹和速度。(2)上层轨迹跟踪控制器设计。提出基于全线控自动轮智行底盘性能的分层式集中轨迹跟踪控制架构与策略,设计基于模型预测控制理论的横纵向耦合轨迹跟踪控制器,该控制器用于计算上层期望车辆动作——期望广义力和力矩。基于安全车速模型,引入轨迹跟踪困难程度的概念,设计期望运动调整策略。在极限工况下,规划动作超过路面极限时,主要通过PID算法调整降速程度,并采用模糊理论自适应调整PI参数,从而保证轨迹跟踪精度和车辆稳定性。(3)轮胎力分配与控制。提出基于稳定裕度的最优轮胎力分配方法,通过各车轮的动态载荷增益和轴静态载荷增益调整车轮权重系数,保障各车轮轮胎利用率的均衡,从而提升车辆的稳定裕度。针对四轮驱动/制动和四轮转向的自动轮智行底盘车辆,构建整车控制量的可行域,并提出优先舍弃纵向目标,以保障横摆稳定性为主的轮胎力分配目标调整策略,确保轮胎力分配层始终存在可行解。对Dugoff轮胎模型求逆,并考虑法向载荷对轮胎刚度的影响,结合车轮旋转动力学模型,实现侧向力和纵向力的精确控制。(4)硬件在环实验验证。搭建基于NI PXIe-8861和Micro Auto Box平台的硬件在环(HIL)仿真实验平台,分别对轨迹规划算法和轨迹跟踪控制算法与策略进行独立验证,最后对两者进行了综合验证。实验结果表明,本文提出的轨迹规划方法在避障能力、轨迹平滑性以及合理性上,优于传统人工势场法(APF)和基于正弦阻抗网络(SRN)的规划方法。在不同车速条件下,轨迹跟踪实验显示,较传统分配控制和横纵向解耦控制,本文方法能在极限工况下保障车辆稳定且有效地跟踪期望轨迹。通过静态和动态障碍物综合避让实验,验证了轨迹规划超出路面附着极限时,主动降速策略有效提高了轨迹跟踪精度,并充分发挥了自动轮智行底盘的性能。更多还原

【Abstract】 As the process of vehicle automation continues to accelerate,the trajectory tracking control performance for traditional vehicle configurations has been fully utilized and developed.The breakthrough of robotic wheel technology has transcended the conventional layout of vehicles,endowing them with new driving modes and expanding their ultimate performance.However,trajectory tracking control techniques based on traditional vehicle configurations struggle to fully exploit the performance advantages of intelligent chassis vehicles with robotic wheels and face difficulties in tracking desired trajectories under extreme driving conditions.This study is based on the National Key Research and Development Program(2021YFB2500703),"Multi System Efficient Integration of Hub Motor Action Module and Vehicle Torque Vector Distribution Technology,"and the Guangxi Major Science and Technology Project(guige AA24206032),"Intelligent Drive-by-Wire Chassis Control Technology for Distributed Four-Wheel-Drive Vehicles."The research focuses on the robotic wheel&intelligent chassis vehicle and investigates nonlinear stability control for tracking the desired trajectory of vehicle motion.The key research tasks are as follows:(1)Vehicle Dynamics Modeling and Desired Trajectory Generation.The research first constructs a dynamics model for a robotic wheel vehicle with four-wheel independent driving/braking and four-wheel independent steering,incorporating the effects of aerodynamic drag.The Dugoff semi-empirical tire model is employed to describe nonlinear tire characteristics,and a rotational dynamics model of the wheels is constructed,along with computations for some key parameters.A trajectory planning method based on a spatio-temporal risk legacy model is proposed.In the Frenet coordinate system,a cubic spline networked road is constructed,and a spatio-temporal legacy risk model is established for the road,main lane,and obstacles to enhance the representation of dynamic environmental risks.Dynamic programming is utilized to compute the desired driving trajectory and speed profile.(2)Design of Upper-Level Trajectory Tracking Controller.A hierarchical centralized trajectory tracking control architecture and strategy is developed based on the performance of the fully wire-controlled robotic wheel&intelligent chassis.A longitudinal-lateral coupled trajectory tracking controller based on model predictive control theory is designed to compute the upper-level desired vehicle actions—desired generalized forces and torques.Based on the safe speed model,the concept of trajectory tracking difficulty is introduced,and a desired motion adjustment strategy is designed.Under extreme conditions,when the planned trajectory exceeds the road adhesion limits,the PID algorithm is mainly used to adjust the degree of speed reduction,and fuzzy theory is used to adaptively adjust the PI parameters,ensuring both trajectory tracking accuracy and vehicle stability.(3)Tire Force Distribution and Control.An optimal tire force distribution method based on stability margin is proposed.The dynamic load gain of each wheel and the static load gain of the axle are used to adjust the wheel weighting coefficients,ensuring the balance of tire utilization for each wheel,and thereby improving the vehicle’s stability margin.For the robotic wheel&intelligent chassis vehicles with four-wheel drive/braking and four-wheel steering,a feasible region for vehicle control variables is established,and a tire force distribution target adjustment strategy that prioritizes the abandonment of longitudinal targets to ensure yaw stability is proposed,ensuring that the tire force distribution layer always has a feasible solution.To achieve precise lateral and longitudinal force control,the inverse of the Dugoff tire model is computed,taking into account the influence of normal load on tire stiffness,and integrated with the rotational dynamics model of the wheels.(4)Hardware-in-the-Loop Experimental Validation.A hardware-in-the-loop(HIL)simulation experimental platform is constructed based on NI PXIe-8861 and Micro Auto Box to independently validate the trajectory planning algorithm and the trajectory tracking control algorithm and strategy,followed by a comprehensive validation of both systems together.The experimental results demonstrate that the trajectory planning method proposed in this paper outperforms the traditional Artificial Potential Field(APF)method and the Sine Resistance Network(SRN)-based planning method in terms of obstacle avoidance capability,trajectory smoothness,and rationality.Under various vehicle speed conditions,the trajectory tracking experiments reveal that compared to traditional allocation-based control and decoupled longitudinal-lateral control,the proposed method in this paper ensures stable and effective tracking of the desired trajectory under extreme conditions.Through comprehensive static and dynamic obstacle avoidance experiments,it is verified that when the trajectory planning exceeds the road adhesion limit,the active speed reduction strategy effectively improves the trajectory tracking accuracy and fully utilizes the performance of the robotic wheel&intelligent chassis vehicles.更多还原