Automatic generation control of diverse energy source-based multiarea power system under deep Q-network-based fuzzy-T2 controller
Prakash Chandra Sahu, R. Baliarsingh, Ramesh Chandra Prusty, Sidhartha Panda
Abstract
The present research article has proposed a robust type-2 fuzzy PID (T2-FPID) controller to obtain automatic generation control (AGC) of a diverse energy source-based multiarea power system under various operating conditions. The AGC aims to make necessary control actions over frequency and tie-line power of the power system. The proposed T2-FPID controller is accessed with triangular membership functions for getting improved performance of fuzzy controllers. The parameters of proposed controller are optimally designed with suggesting a deep Q-network (DQN) algorithm under various operating conditions. The performance of proposed T2-FPID controller is validated over type-I fuzzy PID (T1-FPID) controller and PID controller in response to AGC of the multiarea power system. To justify superiority, the performance of DQN algorithm over particle swarm optimization (PSO) and genetic algorithm (GA) has been synthesized through suitable comparative studies. Performance analysis demands superiority of proposed T2-FPID controller for AGC of multiarea power system. The robustness of proposed T2-FPID controller is investigated under wide variation of system parameters and has been carried out through sensitive study. It has been conferred through various simulation results that proposed DQN-based T2-FPID controller gracefully improves the system performances in regard to the AGC of multiarea power system.