基于PID神经网络的无人机三维航迹控制方法研究  

Research on UAV 3D trajectory control method based on PID neural network

作　　者：杨格[1] 闫建国[1] 

机构地区：[1]西北工业大学自动化学院,陕西西安710129

出　　处：《现代电子技术》2014年第8期46-50,共5页Modern Electronics Technique

基　　金：国家自然科学基金资助项目(60974146)

摘　　要：在研究无人机空中加油航迹控制时,发现二维航迹控制方法难以实现高精度的控制要求。其主要问题是二维航迹控制只能解算出水平面的位置,忽略了高度信息。主要描述了三维航迹控制方法及控制器设计。在二维导航控制的基础上,采用非定高航程推算原理解算出无人机实时位置信息。着重分析航点高度信息算法,飞控系统实时跟踪此计算高度,完成轨迹控制。控制器采用PID神经网络方法,与传统PID控制相比能明显改善控制器性能,响应快,超调小,稳态精度高,能够满足无人机三维航迹控制的飞行要求。In the research of UAV aerial refueling flight path control,it was found that the 2D trajectory control method was difficult to realize the requirements of control precision,because only the horizontal position was solved by the method,but the height information was neglected. The UAV 3D trajectory control method and controller design are elaborated in the paper. Based on the two-dimensional navigation control,the UAV real-time location information can be solved by using dead reckoning （DR）without a fixed height. The algorithm of waypoint height information is analyzed emphatically in this paper,in which flight control system tracks the calculated height in real time to fulfill the trajectory control. PID neural network（PIDNN）method is adopted in controller design. Compared with the traditional PID control,PIDNN can significantly improve the performance of the controller with fast response,small overshoot,high steady-state precision,and can satisfy the flight requirements of UAV 3D tra-jectory control.

关 键 词：三维导航 高度航迹控制 航程推算法 PID神经网络 

分 类 号：TN911-34[电子电信—通信与信息系统] TP391.9[电子电信—信息与通信工程]