Hot strip laminar cooling control based on NNPID controller

摘要

Hot strip coiling temperature is one important parameter of performance index in hot rolled strip, and its control system is highly nonlinear. A new neural network PID (NNPID) controller, which is based on PID by means of neural network's ability of selflearning and adaptive, is presented. The NNPID controller is designed by combining neural network with PID control strategy. Additional momentum method, that is an improved BP algorithm, is used in the neural network is analyzed. It presents the control for the uncertain-information, multi-variable and hot strip laminar cooling control based on the NNPID controller. The results show that the dynamic quality of the system is improved, and NNPID has good adaptability.