State Estimation And Coordinated Control For Distributed Electric Vehicles

Due to the attractive features of the Distributed Electric Vehicles (DEV), includ-ing simple and compact structure, short transmission chain, fast and accurate controlresponse, modular drivetrain design et al., it was widely recognized that DEV can fullyrealize the benefits of Electric Vehicles (EV) and presents an important future develop-ment direction. In order to improve dynamic control performance of DEV, it is essentialto develop a state estimation and traction force coordinated control system. This thesisstudied the architecture of state estimation and traction force coordinated control system,and examined key technologies based on the new features of DEV to improve dynamicperformance of DEV.After analyzing the dynamic characteristics of DEV, this thesis constructed the over-all architecture for state estimation and traction force coordinated control system. In thisnovel architecture, multiple state parameters were estimated and used in the traction forcecoordinated control system, where the longitudinal and lateral dynamic control of DEVwere coordinated and optimized.The key technologies in the proposed architecture were studied in details. Regardingstate estimation, at first, key vehicle and operational parameters such as vehicle mass,road slope and tire vertical load were estimated. Then, longitudinal velocity, sideslipangle, yaw rate and tire lateral force were estimated by using a Unscented Particle Filter(UPF), which merged dynamic model and kinematic model and combined information oftraction torque and wheel rotation speed with sensor data from GPS and INS. Regardingtraction force coordinated control, a hierarchical control system was proposed. In theupper level, desired objectives were designed, which considered the variation of sideslipangle. In the middle level, a traction force coordinated control allocation algorithm wasdeveloped, which addressed issues related to fault tolerant control and TCS intervention.In the lower level, an adaptive motor control was utilized to eliminate the non-zero biasintroduced by motor uncertainties. With the proposed traction force coordinated controlsystem, the vehicle dynamic performance was optimized and enhanced.To validate the efectiveness of the proposed control architecture, a simulation plat-form was developed in CarSim and Simulink, and the vehicle test platform was also devel- oped based on former work. Simulations and experiments show that the proposed systemcan estimate state parameters with acceptable accuracy compared with other methodspublished in the literature, and comprehensively address issues of fault tolerance con-trol, wheel spin in coordinated control, which improve longitudinal and lateral dynamicperformance of DEV, and ensure vehicle safety.