摘要
针对带有不确定性的四旋翼飞行器系统,提出一种滑模控制和神经网络自适应相结合的混合控制方法。该方法在滑模控制的基础上,考虑到实际系统中通常存在建模不精确、参数未知等不确定性,构造RBF神经网络在线逼近系统模型的未知函数,采用Lyapunov方法设计自适应律在线估计神经网络权值和模型未知参数,并通过Lyapunov定理验证了系统的稳定性。仿真结果表明,该方法相对于RBF神经网络的自适应PID控制,具有更短的调节时间、更小的超调量和更好的抗干扰能力,同时在模型参数发生变化的情况下,该控制器的鲁棒性能更强。
In view of the uncertainty about the quad-rotor aerial vehicle system,a new control method is proposed by combining sliding mode control with the adaptive neural network.Considering the existing uncertainties in the actual system,such as inaccurate modeling and unknown parameters,and based on the sliding mode control,we constructed a RBF neural network to on-line approach the unknown functions of the system model,and designed an adaptive law to on-line estimate the weights of the neural network and the unknown parameters using the Lyapunov method.The stability of the system was verified by Lyapunov theorem.The simulation results show that： compared with the adaptive PID controller of the RBF neural network,this controller has a shorter settling time,less overshoot and better resistance to disturbances,and it also has a stronger robustness when the model parameters are changed.
作者
韩业壮
华容
HAN Ye-zhuang HUA Rong(School of Electrical and Electronic Engineering, Shanghai Institute of Technology, Shanghai 201418, China)
出处
《电光与控制》
北大核心
2017年第11期22-27,共6页
Electronics Optics & Control
基金
上海市重点课程建设项目(33210M161019)
关键词
四旋翼飞行器
滑模控制
神经网络
自适应律
quad-rotor aerial vehicle
sliding mode control
neural network
adaptive law
