Implementation of AI Based Power Stabilizer Using Fuzzy and Multilayer Perceptron In MatLab

Systematic investigation of an Automatic Voltage Regulator (AVR) indicates one significant tradeoff in the effectiveness of Excitation System i.e. rapid response with high gain of the AVR induces undesirable damped oscillations in an Electrical power system, which slow down the rotor speed; To overcome this problem, Power system stabilizer (PSS) is used in parallel with excitation system (ES), by injecting extra stabilizing signals to minimize the side effect induced by AVR. The PSS must be self-tuned for adjusting parameters and managing different loading conditions. Therefore, this work is mainly focused on Multilayer Perceptron (MLP) feed-forward neural network and fuzzy logic system controllers to tune and adjust the PSS parameters to achieve better enhancement instability for varying load conditions. In this research work, PSS is designed with different controllers in MATLAB/ Simulink. The development of the PSS is achieved by using different controllers like ProportionIntegrator (PI), Proportion-Integrator-Differentiator (PID) and Artificial Intelligence (AI) based fuzzy and MLP controller. Simulation test results of Voltage and Frequency show the robustness of MLP type PSS as compared to PI, PID, and Fuzzy PSS in terms of minimized overshoot peak value, settling time and rise time for varying loading conditions.