Autonomous flight control system for unmanned helicopter using neural networks

This paper describes methods to develop autonomous flight control systems for UAVs. The unmanned helicopter "RMAX" produced by YAMAHA Motor Co., LTD. is used in this study. It was difficult to develop flight control systems, because the dynamics of the helicopter is nonlinear. An efficient method to design controllers by training neural networks is proposed in this paper. It is easy to use trained neural network together with online training neural networks or adaptive controllers to compensate undesirable effects which are not modeled or sudden changes of the target and environment, therefore the control system can be highly reliable. Results of flight experiments are shown to demonstrate the effectiveness of our approach.