A Comparative Analysis of Fuzzy Logic and Genetic Algorithm Techniques for Load Frequency Control in Multi Area Power System

Abstract

Modern Modern interconnected power systems face continuous challenges in maintaining frequency stability due to dynamic load variations and inter-area power flow disturbances. Automatic Generation Control (AGC) serves as the primary mechanism to restore system frequency and regulate tie-line power exchange in multi-area power networks. Conventional PI controllers, although widely used, suffer from fixed-gain limitations that degrade performance under varying operating conditions, necessitating the adoption of intelligent control strategies. This paper presents the modelling, design, and MATLAB/Simulink-based simulation of a four-area interconnected thermal power system using a Fuzzy Logic Controller (FLC) and a Genetic Algorithm-Optimized PI (GA–PI) Controller. The primary objective is to minimize frequency deviation and tie-line power error following load disturbances in each control area. Simulation results demonstrate that the GA–PI controller outperforms the FLC, achieving 24.6% lower average overshoot (0.49%), 19.1% faster settling time (5.0 s), 18.3% lower rise time (1.70 s), 20% lower peak undershoot (0.0036 p.u.), 25% reduction in tie-line power oscillations (0.009 p.u.), and 66.7% lower steady-state error (0.0001 p.u.). MATLAB verification over a 1000 s simulation with 7744 samples further confirms stable operation with zero overshoot and zero steady-state error. These findings establish the GA–PI controller as a robust, adaptive, and computationally efficient solution for multi-area automatic generation control, providing superior dynamic performance and enhanced frequency stability compared with the fuzzy logic controller.