DESIGN OF INTELLIGENT CONTROL SYSTEM BASED ON PID CONTROLLER AND REINFORCEMENT LEARNING FOR ELECTRIC VEHICLE SPEED CONTROL

• Main Author: Aksanul Amal, Muhammad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/56873
• Institution: Institut Teknologi Bandung

Electric vehicles have become the center of attention, considering that electrification transportation is being heavily campaigned as a solution to the increasing emissions of Greenhouse Gases (GHGs). This makes the development of electric vehicles become a hot issue in academics, including the aspect of speed control because it’s related to energy consumption. Unfortunately, the controller that is widely used, i.e., PID, is only limited to tracking signals, not to compensate disturbances (disturbance rejection). In fact, in the operation of electric vehicles, various disturbances are very likely to occur, such as internal disturbances, i.e., the slope of the track and additional loads. Other disturbances can also be in the form of input and output disturbances in the system. Therefore, we need a control system design that can perform tracking signal and disturbance compensation at the same time. in this study, a control system which is a combination of PID controllers and Reinforcement Learning (RL) was developed. The PID controller is used for the setpoint tracking function by optimizing the accuracy, speed, and stability of the response. Meanwhile, in addition to optimizing PID performance, RL also functions to minimize errors when a disturbance occurs. In this study, RL was also used to synthesize PID parameters. RL method that we used in this study is the Deep Deterministic Policy Gradient (DDPG) because this method has continuous action and state. The PID controller synthesized with RL was able to make the system stable and reached an RMSE value of 9.296, better than the Ziegler-Nichols synthesis result of 9.569. In the PID RL controller for any given noise type, the RMSE value is smaller than the ordinary PID controller. With output disturbance, the RMSE of the PID controller is RL 1.174, while the usual PID controller gets 1.274. ?