面向车载网络的电动汽车主动悬架鲁棒控制策略研究

Active suspension makes suspension keep optimal damping state under different conditions, which can break through the limitation of passive suspension to satisfy the electric vehicle's riding comfort and handling safety at the same time. The electric vehicle active suspension is studied in this project through combining theory analysis, simulation and experimental verification. The dynamical model of distributed driven electric vehicle based on active wheel is derived. And the active suspension is modeled too. The communication characteristics of the in-vehicle network are discussed such as data transfer time-delay, data packet dropout. The model of the whole vehicle and suspension dynamics under the in-vehicle network are established to analyze the effect of communication characteristics on the performance of the vehicle dynamics. And then, the influence factors are represented by uncertainty model and robust predictive control algorithm including non- deterministic communication of the network is investigated. Simulation is made by using the software of MATLAB, ADVISOR and other softwares to make further determination of the parameters of the whole electric vehicle and active suspension model, the key parameters of network communication and controller etc. Finally, the electric vehicle with electromagnetic suspension prototype is developed. The network communication, electric vehicle mathematical model , robust predictive controller are all experimentally verified. The research can improve the electric vehicle’s active suspension performance under the in-vehicle network and promote the development of electric vehicle network control technology.

主动悬架在不同工况下可使车辆处于最优减振状态，能突破被动悬架的局限，可同时满足电动汽车操纵稳定性和行驶平顺性要求。本项目以电动汽车主动悬架为研究对象，采用理论分析、软件仿真和实验验证等相结合的方法开展研究。建立主动轮分布式驱动电动汽车动力学模型、主动悬架动力学模型。再分析车载网络CAN总线的传输时滞、数据丢包率等特性，研究含车载网络的电动汽车整车及主动悬架动力学模型，分析车载网络对车辆动力学性能的影响。针对CAN总线的不利影响因素，采用不确定性描述模型，研究含网络通信不确定性的主动悬架鲁棒预测控制算法。采用MATLAB、ADVISOR等软件进行仿真研究，进一步确定电动汽车整车及主动悬架参数、网络通信及控制器关键参数。研制电动汽车整车及电磁主动悬架样机，开展网络通信、电动汽车数学模型及鲁棒预测控制算法等验证工作。本研究能提高车载网络下电动汽车主动悬架性能，可促进电动汽车整车网络控制技术发展。

主动悬架在不同工况下可使车辆处于最优减振状态，能突破被动悬架的局限，可同时满足电动汽车操纵稳定性和行驶平顺性要求。项目针对含车载网络的电动汽车主动悬架系统，研究主动轮分布式驱动电动汽车动力学模型与主动悬架模型，明晰车载网络的传输时滞、数据丢包率等非线性特性，研究非线性因素对主动悬架系统的影响，研究非线性主动悬架鲁棒控制算法，并通过电动汽车整车及主动悬架样机，开展网络通信及鲁棒控制算法等验证工作。.设计了一种集驱动、减振和制动于一体的新型电动轮系统，路面输入和驱动电机的振动激励输入均影响系统性能，轮内悬架阻尼系数对平顺性影响更为明显，多级隔振型和动力吸振型轮毂电机悬架系统构型相比较，后者表现出更好的减振性能，所设计的主动轮系统在宽车速范围内可以保持良好的减振性能。.CAN通信网络具有明显的非线性特性，与吞吐量、通信冲突率均有相关。利用主成分分析法对通信数据进行降维、检测，再采用支持向量机实现通信异常数据检测，保证了检测准确率，又能提高检测效率。.项目全面考察了时滞、饱和、死区、数据丢包等非线性因素对主动悬架系统的性能影响，研究了最优控制算法、抗饱和控制算法、自适应控制算法、分数阶控制算法以及鲁棒控制算法等，均能实现较好的减振性能，相比较而言，鲁棒控制算法有较大的适应性。.项目设计了基于轮毂电机的电动汽车整车及集成式电磁主动悬架方案，研制出基于轮毂电机的小型样车，改装研制了纯电动汽车，开展了样车调试与实验验证工作，对主动轮电动汽车及其主动悬架数学模型、控制算法进行了实验验证。.本研究有望改善电动汽车主动悬架性能，为提高整车性能提供参考。