摘要
针对6自由度非线性欠驱动的四旋翼无人机飞行姿态易受外部干扰、超调较大、响应较慢等问题,文章基于广义动态模糊神经网络(GD-FNN)设计了一种四旋翼无人机姿态控制系统,并基于Lyapunov函数对系统的稳定性做出了证明。该控制系统结合了GD-FNN的优点,使得系统可同步调整学习时的参数和结构且学习速度快,在高复杂性、不确定性以及存在外部扰动的情况下具备较强的鲁棒性和自适应性。实验结果表明,基于GD-FNN自适应控制系统较传统的PID控制器具有更优越的控制性能,当系统受到外部噪声干扰时,能够快速实现四旋翼无人机高精度和超调小的飞行姿态自适应控制。
To overcome the problems of six-degree-of-freedom nonlinear underactuated quadrotor,such as vulnerable to external interference,large overshoot and slow response,the paper proposed a quadrotor attitude control system is designed based on the generalized dynamic fuzzy neural network(GD-FNN),and the stability of the system is proved based on the Lyapunov function.The control system combines the advantages of GD-FNN,so that the system can synchronously adjust the parameters and structure of learning and learning speed is fast,which makes the quadrotor achieve strong robustness and adaptability in the case of high complexity,uncertainty and external disturbance.Experimental results show that the GD-FNN adaptive system designed in this paper has better control performance than the traditional PID controller.When the system is interfered by external noise,it can also quickly realize the high-precision and small-overshoot flight attitude adaptive control of quadrotor.
作者
潘捷
牛萍娟
张伟龙
韩抒真
毛润
PAN Jie;NIU Ping-juan;ZHANG Wei-long;HAN Shu-zhen;MAO Run(School of Electronics and Information Engineering,Tianjin Polytechnic University,Tianjin 300380,China;Information Center,Tianjin Polytechnic University,Tianjin 300380,China;State Grid Tianjin Electric Power Company Chengnan Power Supply Branch,Tianjin 300201,China)
出处
《组合机床与自动化加工技术》
北大核心
2020年第6期123-126,136,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国网天津市电力公司2019年科技服务项目(SGTJCN00YJJS1900527)
天津市科技计划资助项目(18ZXZNGX00130)
国家自然科学基金(11204211)。