摘要
为了提高三级倒立摆系统控制的响应速度和稳定性,在设计Mamdani型模糊推理规则控制器控制倒立摆系统稳定的基础上,设计了一种更有效率的基于Sugeno型模糊推理规则的模糊神经网络控制器。该控制器使用BP神经网络和最小二乘法的混合算法进行参数训练,能够准确归纳输入输出量的模糊隶属度函数和模糊逻辑规则。通过与Mamdani型控制器的仿真对比,表明该Sugeno型模糊神经网络控制器对三级倒立摆系统的控制具有良好的稳定性和快速性,以及较高的控制精度。
In order to improve the response and stability of a Triple Inverted-Pendulum system, a Sugeno fuzzy neural network controller is proposed. A BP network with Least Square Method is used to train the Sugeno fuzzy neural network for theadjustment of the membership functions and fuzzy logic rules of the input and output variables. The results of simulation of the Sugeno fuzzy neural network controller, as compared with the Mamdani fuzzy controller are also given.
出处
《自动化技术与应用》
2008年第11期7-10,78,共5页
Techniques of Automation and Applications
关键词
三级倒立摆
模糊控制
模糊神经网络控制
triple inverted-pendulum
fuzzy control
fuzzy neural network control
