STUDY ON A NOVEL SPEED CONTROL STRATEGY OF HYDRAULIC TURBINES APPLYING ARTIFICIAL NEURAL NETWORK AND METAHEURISTIC OPTIMIZATION TECHNIQUE

This paper proposes a novel method applying artificial intelligence to design an effective control scheme to maintain the speed of hydraulic turbines in hydropower plant at a nominal value corresponding to the defined net frequency. The structure of the proposed control topology comprises of two phases: an artificial neural network – based model reference adaptive controller and a metaheuristic optimization technique. The novel contribution of this work is to create two gain factors which should be determined effectively by a bio-inspired optimization technique to be integrated with the neural network – based controller. Such two factors are completely able to further improve control quality of the system applying the artificial neural network controller. The proposed method is applied to the speed control of a hydropower plant to maintain the net frequency at the nominal value of 50 Hz against random load changes. The simulation results when compared to the conventional regulators such as PID demonstrate the superiority of the proposed control methodology.