模拟电路实现的神经元控制器的仿真研究  

Simulation Research on a Kind of Analog Neuron Adaptive Controller

作　　者：罗文广[1] 

机构地区：[1]广西工学院电子信息与控制工程系,545006

出　　处：《计算机仿真》2001年第6期39-41,共3页Computer Simulation

基　　金：广西自然科学基金 (编号 :桂科自 0 0 660 0 5 );广西工学院科研基金资助

摘　　要：该文研究一种适用于电动机控制的模拟电路实现的神经元自适应控制器。根据神经元的特性 ,将数字神经元控制器模拟化 ,获得模拟神经元控制器。该文还研究用MATLAB中的动态仿真工具SIMULINK对其进行仿真的方法。仿真结果表明模拟电路实现的神经元控制器比模拟PID具有更良好控制特性 ,并实现神经元权值的自动调节。该仿真方法为用电路仿真软件进行电路设计、仿真以及实际电路的实现打下良好的基础 ,提高设计效率。This paper discusses a realization method of a kind of analog circuit realization neuron adaptive controller applying to motor control. According to neuron characteristic, an analog neuron controller is obtained through simulating digital controller. A simulation approach with SIMULINK is also studied. Simulating results have showed this controller has more well control performance than analog PID controller, and its neuron weights can auto-modulate. The simulation approach provides a good basis of designing and simulating the controller with circuit simulation software, and manufacturing the actual circuit. It also promotes the designing efficiency.

关 键 词：模拟电路 神经元控制器 仿真 电动机 自适应控制器 

分 类 号：TM32[电气工程—电机] TN710[电子电信—电路与系统]