摘要
给定动态系统的控制主要解决三个问题:动态系统的稳定性、控制目标以及干扰控制。传统的控制方法使用反馈同时解决这三个问题,需要折中获得较好的效果。本文提供一种分开解决这三个问题的方法。这种方法适用于单输入单输出和多输入多输出的线性系统以及非线性系统。通过在输出端获取干扰信号,经过滤波,反馈到输入端来完成干扰的最优控制,且反馈回路并不影响系统的动态响应。以被控对象的逆作为控制器,可以准确跟踪系统的输入。最后把自适应逆控制器用于无人机的着陆以验证该控制器的有效性。
For controlling a given dynamic system, there are three problems need to be solved: stability of the dynamic system, control goal and disturbance control. The traditional method to solve these problems simultaneously is feedback, which needs compromise to achieve better result. This paper proposes a method to solve the problems separately, which can be used to handle linear or non-linear SISO and MIMO systems. The disturbance signal is obtained at output terminal, filtered and then feedbacked to input terminal to accomplish optimal control, and the feedback circuit can' t affect dynamic response of the system. Taking the inverse of the plant as controller, input signal is tracked precisely. This method to control the landing of UAV with adaptive inverse controller is effective by test.
出处
《航空兵器》
2009年第4期3-6,11,共5页
Aero Weaponry
基金
航空支持基金(05C52007)
关键词
自适应逆
神经网络
无人机
着陆
adaptive inverse control
neural network
UAV
landing