Identification and PID Control of ASTANK2 Plant Using Genetic and Firefly Algorithms

The paper presents the application of metaheuristic optimization techniques in identification and control of ASTANK2 installation. The optimal MIMO identification model is found by solving the problem of Signal-to-Noise Ratio maximization, with the help of two metaheuristics: the Firefly Algorithm and the Genetic Algorithm. Subsequently, a PID-based control strategy of ASTANK2 is adopted. The simulation results have shown that the Genetic Algorithm performs slightly better than the Firefly Algorithm in model identification. Therefore, the optimal parameters of PID controllers that ensure the desired behavior of installation are tuned by using the Genetic Algorithm. A robustness analysis is developed, concerning the performances of controllers to output disturbances rejection and staircase references tracking.