过驱动多轮直驱轻型电动汽车多模型估计及控制研究

The multi-wheel independently-actuated electric vehicle with in-wheel motors which is lightweight and miniaturized has become a focus of research on a worldwide scale. Due to the large reductions in vehicle mass and body size, it has been shown that the vehicle dynamics are more sensitive to payload parametric uncertainties/variation,i.e.,passengers or freight loads. The lightweight electric vehicle (LWEV) with four or two wheel independent steering/driving/braking is the classic over-actuated system with actuation-redundancy. It is obviously important to study actuated states of in-wheel motors, parameters sensitivity, state parameters of over-actuated LWEV for multiple model estimation and control..The electric vehicle(EV) system high order nonlinear dynamics models with multi-link independent suspension are built, and different types of uncertainties and time-varying characteristics based vehicle dynamics configuration and state parameters are shined upon to the model set. Considering vehicle variables (such as lateral velocity, friction forces and tire-road adhesion coefficient) can not be measured directly in modern vehicle, and the interacting multiple model filter estimators with multiple nonlinear dynamics models and multiple parallel nonlinear filters is investigated. Furthermore, payload variation of LWEVs with in-wheel motors affecting manipulating and control performance stability are also analysed, and the active coordination control with independent steering/driving/braking electric vehicle and the dynamic control allocation methods of coupling compensation for multi-actuators are studied with systematic utilization of actuation-redundancy to achieve output independent parameter estimation and its seamless synthesis with adaptive control for improving transient control performance of over-actuated LWEV systems under extreme conditions.

质量轻、小型化的轮毂电机独立驱动电动汽车已成为国内外研究热点，新型专用底盘的设计、质量的重新分布和车身尺寸大幅减小，外界干扰或载荷变化对整车动力学影响较大，融合多轮（四轮或两轮）独立转向/驱动/制动等特性的轻型直驱电动汽车是典型的过驱动系统，研究轮毂电机的驱动状态、车辆参数灵敏度及参数摄动、多模型估计及控制显得尤为重要。.首先建立独立电动多轮驱动多连杆悬架底盘系统的非线性动力学模型，将车辆结构参数、状态变量等具有多种不确定性的时变车辆参数空间映射为模型集；然后考虑侧向车速等状态信息及参变量很难直接测量，设计非线性交互式多模型估计器，使得每个模型的滤波器并行工作，对各模型滤波器所做的估计进行数据融合；分析车辆载荷参数变化（质量、惯性力矩、质心位置等）的灵敏度；考虑多轮驱动电动汽车多执行器的特点，提出多模型车辆高容错协调主动控制分配方法，实现多模型协调主动控制分配的柔顺切换，提高车辆控制性能。

本项目以安全节能的轻型直驱纯电动汽车为研究对象，建立独立电动多轮（四轮或两轮）驱动多连杆悬架底盘系统的非线性动力学模型，研究融合多轮独立转向/驱动/制动等特性的过驱动车辆系统动力学物理特性，设计基于交互式多模型的高速运动车辆系统状态估计；分析轻型轮毂电机驱动电动汽车载荷参数变化（质量、横摆惯性力矩、质心位置等）的灵敏度；考虑多轮驱动电动汽车多执行器的特点，建立多模型车辆高容错协调主动控制分配方法，实现输出独立的状态参数估计并无缝集成自适应控制，提高极限工况的车辆系统瞬态控制性能。. 本项目超额完成预期研究成果，成功构造具有多种不确定性特征的轻型直驱电动汽车系统动力学模型集；设计过驱动多轮直驱轻型电动汽车系统交互式多模型估计算法的理论并建立多模型控制系统估计与协调主动控制理论框架，为广义的具有过驱动特征的车辆、飞行器或船舶系统提供理论支撑；通过开放式硬件在环仿真和实车试验平台实现上述相关技术的验证。. 本项目的完成推动了电动汽车安全节能的发展，为车辆在各种工况下的安全、平顺、可靠行驶提供理论支持与依据。此外，项目涉及的多模型估计及控制方法，对航天 、机器人和生物医学系统等其他过驱动系统的控制也有实际应用价值。