The development of wind velocity adaptive control system based on neural networks for NF-3 wind tunnel

This paper describes the structure and the characters of wind velocity control system for NF-3 low speed wind tunnel which is located at Northwestern Polytechnical University.Due to the difficulties of establishing precise mathematical model,the complicated nonlinear characteristics of the system and the change of parameters of the wind tunnel itself,the velocity system is not easy to control in the traditional control theory.In this new system we applied the multithreading digital data acquiring method,PSD neural networks with dead line control algorithm to solve these problems.Compared with the former fuzzy control systems,this system is not only easier to operate,less time to stable,but also have higher control performance from 0.3% to 0.1% when the velocity is greater than 10m/s.