Model Reference Adaptive Dynamic Inversion Flight Control Based On Neural Networks

The purpose of flight control is to ensure the stability and maneuverability of the aircraft duringits flight. The technology of flight control, as the core part of the flight control, in recent years hasbeen widely used in the field of aerospace and become the hot researches in flight control field.In the thesis, the flight control theory based on adaptive inverse is investigated. At first, itintroduces the main contents, the development history and research status at home and abroad. On thisbasis, an adaptive inverse flight control program based on BP neural network for a civilian aircraft isproposed and analyzed. Moreover, ant colony algorithm is applied to optimize the BP neural networkweights, which avoids the local optimal and improves the convergence rate. The forward and sidedirection flight performance simulation of the original algorithm and improved algorithm is be done,the result shows that the improved flight control program could effectively compensate for the inversemodeling errors, and obtains good robustness. The new flight control strategy can improve theperformance of the flight control system so that pilot can complete the mission safely.At last, the cloud flight simulation platform is developed based on Java language and Matlabwithout client installation and based on WEB, which could be simulated where there is Internet.Having overcome the shortcoming of traditional simulation software based on local, the simulationresults show the effectiveness of the simulation platform.