本論文主旨在於發展一套馬達驅動式電動汽車之運動控制系統。此系統包含內外兩層控制迴路。外迴路以直接搖擺力矩控制法為基礎,負責掌控車輛轉彎時的搖擺角速度以增進轉向之穩定性,而內迴路則由參考模式適應性控制演算法為主幹建構;此演算法是根據李亞普諾夫穩定定理來即時計算控制器之參數,可有效提升車體之強健性。除此之外,最佳電流波型控制法也應用於本研究中來提升馬達之效率。 在實驗部份,直接搖擺力矩控制法建立於MATLAB®/Simulink®軟體環境中並藉由電腦執行,而電流波型控制與參考模式適應性控制演算法則以Verilog硬體語言配合定點數運算撰寫程式以FPGA實現,如此可大幅加快程式執行速度與節省硬體資源。最後使用經過改裝的Nissan March搭載由本實驗室開發之輪內馬達,以後輪驅動之型式進行實車測試來驗證本文所提出之控制方法的性能。
This thesis proposed an electrical vehicle motion control system including two parts: (1) Lateral vehicle motion control: An outer-loop direct yaw moment control (DYC) scheme is established to manage the yaw motion of vehicle during cornering. (2) Longitudinal vehicle motion control: An inner-loop model reference adaptive control (MRAC) base on the Lyapunov theory is applied to compensate the speed response of the two driving wheels for the improvement longitudinal vehicle motion and gain the system robustness. Furthermore, the optimal phase current waveform control is applied in this thesis to improve the efficiency of motor. In experiments, the inner-loop DYC is constructed with MATLAB®/Simulink® software and performed through a PC; at the same time, the current control and inner-loop MRAC are implemented with an EPF10KE Field Programmable Gate Array (FPGA) coded by using Verilog hardware description language (Verilog HDL) in fixed-point arithmetic format so as to relieve the complex logic circuit synthesis in FPGA and time-consuming computations. Finally, the advantages of the proposed methodologies is verified by employing the commercial vehicle “Nissan March”, which is reconstructed to be driven directly by two in-wheel motors.