Modeling And Dynamics Control Of Four-wheel-independent-drive Electric Vehicle With Network-induced Delays

In the transportation field, the electric vehicle(EV) is considered as an effective solution for the energy and enviromental issue. A special type of EV named four-wheel-independent-drive electric vehicle(4WID-EV) with in-wheel motors has gain consistent attention from both automobile industry and academia. On the other hand, the development of X-by-Wire technology and in-vehicle network has promoted the application of distributed control in vehicle mechatronic systems. The influence of network communication issue on the control performance has also been widely studied in the control theory field. Moreover, it is being concerned in the engineering and application field. Taking the 4WID-EV with distributed control scheme as research object, this dissertation focuses on the influence of in-vehcile network communication delays on the vehicle dynamics control, and conducts the research of system modeling, control algorithm design and validation.Starting with the stablity evaluation index of control system, fundamental factors in the distributed control with their impact are studied. The influence and relationship of delay length, control paramters and accuracy, delay distribution character and other factors are studies in detail by means of simulation, with the slip ratio control of electric wheel as typical application. For further control algorithm design of 4WID-EV dynamics, a control-oriented time-lag dynamic model with time-varying random delays is built.Aiming at 4WID-EV with active front-wheel steering(AFS) system, an upper-level controller which is robust to time-varying random delays and an optimal torque allcoation algorithm with fault-tolerant capability for communication interrupt are proposed in the hierarchical control architecture. With the time-lag vehcile dynamic model, the matrix polytope model is adopted to describe the system uncertainties induced by the time-varying random delays, upper-level controller is designed with the H∞ robust theory and LQR, and linear matrix inequality(LMI) is used to solve the optimal feedback gain of the controller. Based on four-wheel rigid-body vehicle model and magic formula tire model, the control allocation theory of over-actuated system is employed to design the optimal torque allocation algorithm in the form of quadratic programming problem. The proposed allocation algorithm can calculate the driving/braking torque of the four wheels to satisfy the direct yaw-moment desired by the upper-level controller, and meanwhile optimize the control variables of the actuators and the friction between the tire and road. Considering the communication interrupt between the bottom motor controllers and in-vehicle network, a fault-tolerant mechanism is designed within the frame of optimal torque allcoation. Co-simulations with Matlab/Simulink and Car Sim software verify the effectiveness of the upper-level robust controller and optimal torque allocation algorithm, respectively, as well as the robustness to time-varying network delays and fault-tolerant capability for communication interrupt. In addition, operating condition-based dynamic message priority scheduling mechanism is proposed. The effectiveness of the mechanism is preliminarily validated with simplified network model, and the bottom-level implementation scheme is briefly discussed with in-vehicle network protocol.A hardware-in-the-loop(HIL) platform for vehicle distributed control system is constructed to further verify the proposed control algorithm. The HIL platform is made up of vehcile control unit with MPC5644 A microcontroller, realtime vehicle model based on CarSim RT and MicroAutoBox, CAN bus supervisory control module with VN7600 and CANoe, upper computer, etc. The feasibility, effectiveness, robustness against time-varying delays, and fault-tolerant capability for communication interrupt are validated for the hierarchical control architecture with proposed control algorithm.