基于自适应神经网络Backstepping空中加油编队飞行控制  

Flight Control of Air Refueling Formation Based on Adaptive Neural Network Backstepping

作　　者：李华东 袁锁中[1] 安帅 Li Huadong;Yuan Suozhong;An Shuai(College of Automation Engineering,Nanjing University of Aeronautics ＆ Astronautics,Nanjing 210016,China)

机构地区：[1]南京航空航天大学自动化学院,南京210016

出　　处：《兵工自动化》2018年第11期1-4,11,共5页Ordnance Industry Automation

摘　　要：为解决无人机自主空中加油过程加油机和受油机编队飞行控制的问题,提出了对受油机采用导引回路和姿态回路分离控制策略,采用L1导引算法设计一种基于自适应神经网络Backstepping飞行控制器。将改进的L1导引算法应用于受油机的横向和纵向的导引,使用自适应神经网络补偿受油机受到的外界的干扰和系统模型误差,神经网络权值矩阵通过自适应律在线更新,并结合Backstepping控制方法设计受油机的控制律。仿真结果表明:在受油机与加油机编队飞行过程中,该设计方法能有效提高受油机的跟踪精度和抗干扰能力,解决空中加油编队飞行控制问题。In order to solve the process of UVA autonomous aerial refueling tankers and receiver aircraft formation flight control problem, proposed to the receiver aircraft separation by guidance loop and attitude control strategy, and design based on the L1 guidance algorithm based on adaptive Backstepping flight controller. Improved L1 guidance algorithm was applied to horizontal and vertical guided by oil machine, using adaptive neural network compensating the receiver aircraft external disturbance and system model error, the neural network weight matrix through online updates, the adaptive law and combining the Backstepping control method to design the receiver aircraft control law. The simulation results show that the receiver aircraft and tankers in the process of the formation flight, the design method can effectively improve the tracking precision of the receiver aircraft and anti-interference ability, effectively solve the problem of aerial refueling formation flight control.

关 键 词：自主空中加油 编队飞行 L1导引 自适应神经网络 BACKSTEPPING 

分 类 号：TP202[自动化与计算机技术—检测技术与自动化装置]