Comparative Performance of UPQC Control System Based on PI-GWO, Fractional Order Controllers, and Reinforcement Learning Agent

In this paper, based on a benchmark the performance of Unified Power Quality Conditioner (UPQC), improvement is presented comparatively by using Proportional Integrator (PI)-type controllers optimized Grey Wolf Optimization (GWO) computational intelligence method, fractional order (FO)-type differential and integral calculus, PI-type controller in tandem with Reinforcement Learning—Twin-Delayed Deep Deterministic Policy Gradient (RL-TD3) agent. The main components UPQC are series active filter an Active Parallel Filter (APF) coupled to common DC intermediate circuit. provides voltage reference for APF, which turn corrects both harmonic content introduced load VDC improved types listed above APF structure. indicators UPQC-APF control system are: stationary error, ripple, fractal dimension (DF) Results also current Total distortion (THD) case of, respectively, linear nonlinear highly polluting terms content. Numerical simulations performed MATLAB/Simulink environment demonstrate superior when PI RL-TD3 agent FO-type compared classical controllers.