基于BP神经网络动态逆的纵列式直升机控制  

Tandem helicopter control based on BP neural network dynamic inversion

作　　者：张翾 吴梅[1] 王宇航 ZHANG Xuan;WU Mei;WANG Yuhang(School of Automation,Northwestern Polytechnical University,Xi’an 710129,China)

机构地区：[1]西北工业大学自动化学院,陕西西安710129

出　　处：《飞行力学》2021年第6期36-41,共6页Flight Dynamics

摘　　要：针对纵列式直升机存在的强耦合、强非线性、旋翼之间气动干扰大及非线性模型难以求解问题,设计了基于BP神经网络的动态逆控制器。采用线性动态逆简化了控制问题,在此基础上引入在线BP神经网络对线性动态逆系统,补偿由非线性建模和小扰动线性化所导致的逆系统误差。仿真结果表明,所设计的控制器能够有效降低系统超调量、减小系统响应时间,说明控制器具有良好的动态性能,适用于实时控制的场合。A dynamic inverse controller based on BP neural network was designed for the tandem helicopter to solve the problems of strong coupling,strong nonlinearity,large aerodynamic interference between rotors,and difficulty in solving nonlinear model.The linear dynamic inverse was used to simplify the control problem,and the on-line BP neural network was introduced to the linear dynamic inverse system to compensate the error of the inverse system caused by nonlinear modeling and small disturbance linearization.Simulation results show that the designed controller can effectively reduce the overshot and response time of the system,which shows that the controller has good dynamic performance,and is suitable for real-time control occasions.

关 键 词：BP神经网络 动态逆 误差补偿 伪控制输入 模型控制输入 

分 类 号：V249.1[航空宇航科学与技术—飞行器设计] V275.1