Reinforcement Learning Based Autonomous E-Vehicle Speed Control

Greater efficiency in both energy use and traffic flow are two benefits of autonomous self-driving cars. Due to their superior performance, effectiveness, and lack of carbon emission, electric vehicles (EVs) have recently become popular and used in an autonomous vehicle. As EVs are making a splash in the automotive industry, researchers are taking a greater interest in studying, modelling, and simulating them. Controlling EV speed is not an easy task. Modelling and Conventional Proportional Integral Derivative (PID) controller tuning for EV speed regulation are presented in this paper. To design and simulation of EV speed control in MATLAB, the transfer function of EV is derived and considered. PID controller is found to be easy to implement, practical, and provide superior closed-loop performance. Two PID tuning techniques, Ziegler-Nichols (ZN) and a Reinforcement Learning (RL) method are designed to regulate the EV speed. Characteristics of the time domain have been used to perform a comparative analysis. Additionally, the Integral Square Error (ISE) is analysed to determine the optimal PID tuning strategy for EV speed regulation.