Efficacy of system optimization by Genetic Algorithm with and without noise

摘要

The paper presents two plants which use proportional-integral-derivative (PID) controller for obtaining the desired transient response. Ziegler Nichols (ZN) and Genetic Algorithm (GA) tuning methods are employed on both the plants affected by band-limited white noise, and the optimized parameters (viz. Settling Time and Peak Overshoot) are compared. The efficacy of system optimization by GA is being tested so that both set point tracking and disturbance rejection are enabled, hence posing a servo as well as regulator problem. The result shows that GA proves to be more favorable for optimization of systems with or without noise as compared to ZN tuning method.