Pid Control Parameters Based On Bp Neural Network Self-learning In The Nf-6 Wind Tunnel Model Attitude Control

NF-6 Wind tunnel is a pressurizing, continueus, high-speed and return-flow aerofoil wind tunnel. It is the first Pressurizing, continuous and transonic wind tunnel in our country, and it is the world's first class tremendous high-speed aerofoil wind tunnel with the high performance claim for control system and conventional control strategy can't meet with the request. A type of self-adjusted PID control strategy based on BP NN is given in this paper. In order to improve the algorithm astringency, adopt the multiple derivative back propagation algorithm, and there is farther analysis and improvement combining with improved PID algorithm, in a certain extent knock-down dependence on exact model and solve the question of online adjust. The astringency analysis shows that with the suitable parameters and in some certain range, this algorithm is perfectly convergent. Experiments result that the control strategy in this paper is with high disturbance rejection performance and it's able to bring on a perfect control effect.The wind tunnel mode attitude system software is the hard core of the whole system,and it constitutes the integrate work environment. This system adopts LabWindows/CVI, the American National Instrument corporation's virtual instrument developing software, develop control system software platform of model attitude.This paper also describes network communication of wind tunnel control system.