基于模糊神经网络的电动执行机构

Research of Fuzzy Neural Network Controller Application in Electric Actuator

电动执行机构控制的对象往往有多参数、非线性、时变以及变量强耦合的特点,很难建立精确的数学模型。模糊神经网络控制系统利用神经网络的非线性映射能力完成模糊控制,能很好地解决控制对象的动态特性所具有的非线性、时变性、参数可变等问题。仿真对比试验表明,电动执行机构采用模糊神经网络控制器(FNNC)后,系统的响应速度变快,调节精度提高。该控制器的适应性、鲁棒性也明显优于常规P ID控制器。

The object of electric actuator is usually a multi--parameter, nonlinear, time varying and strongly coupling course, so the exactly mathematical models are hardly constructed. The fuzzy neural network control system, which carries out fuzzy control through the capability of neural network^s nonlinear mapping, can solve these problems better. Simulation results show, the electric actuator which is applied with fuzzy neural network controller, improves response speed and adjusting accuracy. The adoption and robust of this controller is obviously better than common PID.