Skill-based PID control by using neural networks

This paper reports a new PID control scheme by using neural networks. This control method can enable a plant to operate smoothly as the plant condition varies with any unexpected accidents. As taking a concrete subject of investigation, we choose a temperature control for a chemical plant which produces polythlene in a batch reactor. Here, the reaction in a tank is complex and has many nonlinear factors. The temperature control is realized by valves manipulation. Since the valve control have been performed by using a conventional PID controller, it generally needs much effort and time to tune the PID gains. Thus in the production field, a new control system with a cheep and simple structure is needed to tune PID gains without operator's decision under various conditions of the plant. For that purpose, we use the self-tuning neuro-PID control method which has the characteristic of tuning PID gains automatically by neural networks. From the experiment results, we show the effectiveness of the proposal method to improve the control performance of the temperature in the batch reactor.