The development of wind velocity control system based on PSD neural networks

This paper describes the structure and characters of wind velocity control system for the NF-3 low speed wind tunnel which is located at Northwestern Polytechnical University in China, due to the difficulties of establishing mathematical model, the complicated nonlinear characters and the change of parameters when time varying, the system is not easy to control with classical algorithm. The new system presented in this paper applies the multithreading digital signal acquiring method, PSD neural networks algorithm, compared with the former fuzzy control system , the later is not only easier to operate, consuming less time to be stable, but also has higher control performance up to 0.1%.