基于开放平台和神经网络的自主飞行控制系统研究

Research on the Autonomous Flight Control System Based on Open Platform and Neural Network

开发了飞行控制软件的应用中间件,实现控制结构的动态变化和控制组件的动态重构. 基于模型逆技术,设计了某状态下的旋转逆和平移逆控制器,其它状态下的逆误差由在线神经网络补偿.实例仿真的结果表明:开放平台和神经网络的结合,能够很好地实现无人直升机的自主飞行,具有较大的理论意义和工程应用价值.

The application middle ware of flight control software is developed, to implement the dynamic change of control architecture and the dynamic reconfiguration of control components. Based on model inversion technology, the rotation inversion and translation inversion controllers under some condition are designed, and inversion errors under other conditions are compensated by the on line neural network. The example simulation results demonstrate that the method of the open platform combined with neural network can implement the autonomous flight of unmanned helicopter and is of great value both in theory and in engineering application.