Application of neural network based PID method for temperature control of aquaculture greenhouse

Abstract. The aquaculture greenhouse is a very complicated system with non-linearity, large delay and time variant due to high specific heat capacity of water. The traditional PID control effect is often not satisfactory for the temperature control of aquaculture greenhouse. In this paper, a neural network based PID control strategy is proposed to regulate the temperature, and a corresponding controller is designed. In this controller, the PID control parameter can be tuned real-timely by neural network. The traditional PID control and the neural network based PID control are analyzed by using MATLAB. Simulation results show that the neural network based PID control strategy has quick dynamic response, small overshoot, strong robustness and better control effect.