基于控制分配的车辆底盘协调控制关键技术研究

To solve the over-actuated vehicle chassis control problem with control objectives of longitudinal and lateral motion and with actuators including front/rear steering and four wheel independent drive/brake, a control allocation (CA) based modular reconfigurable coordination control approach is introduced in this proposal. To be robust to parameter variation of the system model, an adaptive Multi-input Multi-output (MIMO) sliding mode control algorithm is proposed to calculate the desired global control force/torque in the upper controller. Then a CA approach is introduced to allocate the global control forces to each tyre. The CA algorithm is designed taking account to both control error and real-time performance considering state of the vehicle and actuators. More over, based on the physical constraints, actuator faults, vehicle stability and nonlinear tyre force chareteristics, the coordination rules are proposed to adjust parameters of the CA algrithm on-line to improve performance of the control system and achieve fault-tolerant control. Each of the control modules is realised in Matlab/Simulink environment for numerical simulation of the control system, and a hardware-in-the-loop simulation platform for the proposed coordinated chassis control is also built to verify the control strategy and algorithm in typical maneuvering condition. The proposed research will contribute to solving key technologies on the optimal allocation for the tyre force in the hierachical coordinated chassis control, and is of great significance to the research of the engineering application of the fault-tolerant control algorithm for over-actuated systems.

本研究项目针对以纵向和横向运动控制为目标的可前/后轮主动转向和四轮独立制动/驱动的车辆底盘过驱动控制问题，提出基于CA算法的模块化可重构协调控制方法。为满足对系统模型参数变化的鲁棒性，上层计算理想车身控制力/力矩的控制器用自适应MIMO滑模控制，然后采用CA算法将总控制力/力矩分配给各个轮胎力。根据车辆和执行器状态设计兼顾控制误差和实时性要求的CA求解算法，并基于车辆稳定性程度、轮胎非线性特性、执行器物理约束和失效程度，提出协调规则在线调整CA算法参数以提高控制系统性能并实现容错控制。将各控制模块在Matlab/Simulink软件中实现以建立控制系统数字仿真平台，并构建该底盘协调控制系统硬件在环仿真试验平台，通过仿真试验考察项目提出的控制策略和算法在典型操纵工况下的性能。本项目研究可解决底盘模块化分层协调控制中轮胎力优化分配的关键技术问题，且对过驱动系统容错控制算法的工程应用有重大意义

本研究项目针对以纵向和横向运动控制为目标的可主动转向和四轮独立制动/驱动的车辆底盘过驱动控制问题，提出了基于CA算法的模块化可重构协调控制方法。为满足对系统模型参数变化的鲁棒性，上层计算理想车身控制力/力矩的控制器用自适应滑模控制，在线调节滑模控制器增益值，增强滑模控制器对模型误差和外部干扰的鲁棒性；采用CA算法将总控制力/力矩分配给各个轮胎力。主动转向控制和主动驱动/制动控制的权重，在轮胎力优化分配算法中横摆力矩跟踪和纵向速度跟踪误差的权重，以及轮胎力控制输出大小的权重值可以依据车辆稳定性因子进行在线调整。将各控制模块在Matlab/Simulink 软件中实现以建立控制系统数字仿真平台，并且结合CarSim中的汽车模型和Matlab/Simulink中的控制算法进行混合仿真研究。本项目还考虑轮胎附着极限，进行路面附着系数等车辆参数估计，以及在轮胎力附着极限工况下四轮独立驱动汽车轮胎力的解析分配算法，并通过LMI方法求解。本项目成功改装了一台性能优越的四轮独立驱动+主动转向电动实验车，该实验车可以进行前轮主动转向和纵向轮胎驱动力优化分配算法的实车试验。本项目研究解决了底盘模块化分层协调控制中轮胎力优化分配的一些关键技术问题，对过驱动系统容错控制算法的工程应用有重大意义。