Motor Torque Estimation and Security Control for Electric Vehicle Applications

Abstract

Electric vehicle (EV) propulsion systems require high-performance control strategies to ensure accurate speed regulation and robustness against external disturbances. Among the factors affecting drive performance, load torque variations caused by road conditions, vehicle mass changes, and resistive forces can significantly degrade speed tracking and dynamic response. This thesis investigates the use of observer-based load torque estimation techniques to enhance the performance of a DC motor drive intended for electric vehicle applications. Three estimation methods are studied: the Luenberger Observer (LO), Sliding Mode Observer (SMO), and Extended State Observer (ESO). Their mathematical formulations are developed and their estimation capabilities are first evaluated under open-loop conditions. The proposed observers are then integrated into a closed-loop speed control system, and their performance is assessed through simulation under various operating conditions. Finally, the developed approach is applied to a 1.5 kW DC motor-based electric vehicle propulsion system. The results demonstrate that accurate load torque estimation improves disturbance rejection, speed tracking performance, and overall system robustness.