模糊神经网络控制器的设计与仿真

The Design and Simulation of the Control Device of Fuzzy Nerve Network

提出了一种复杂系统模糊建模及控制的神经网络方法。通过BP神经网络，对模糊规则的结论参数和隶属函数进行在线修正，实现模糊规则的自组织。这种控制方法可用于对受控对象缺乏精确的数学模型或具有时滞、高阶、非线性等难以用现有的控制理论方法分析和控制的复杂系统。仿真结果证明了这种方法的有效性。

The article puts forward a complex system for fuzzy model-building and a method for a control nerve network .By means of BP nerve netWork, the conclusive parameters of the fuzzy principles and subordinate function are revised on line and self-organization of the fuzzy principles is realized. This control system can be used for a controlled object which lacks an accurate mathematical model or for a complex system which is difficult to analyze and control with the existing control method because of time-delay and nonlinearity. The simulation results show the effectiveness of this method.