摘要
新建一个3层BP神经网络的虚拟仪器程序,通过设置隐含层层数、学习速率、惯性系数、输入信号类型和采样周期等控制单元,可将PID控制器的调控参数传递给后台程序.程序运行时,用户在控制面板中修改参数的输入值,系统做出实时响应,系统输出波形也会发生相应的变化.仿真结果表明,BP神经网络PID控制器参数调整简单,具有较高的精度和较强的适应性.
A new layer 3 BP neural network of virtual instrument in the program is created.By setting the implicit layers,learning rate,inertia coefficient,the input signal types and sampling period control unit,etc.,the PID controller parameters can be passed to the backend application.When the program is running if parameters of the input value are modified in the control panel,the system will make real-time response,at the same time the output waveform of the system will change.Simulation results show that the adjustment method of BP neural network PID controller parameters is simple and it has higher precision and strong adaptability.
出处
《华侨大学学报（自然科学版）》
CAS
北大核心
2011年第4期475-477,共3页
Journal of Huaqiao University(Natural Science)
基金
福建省自然科学基金计划资助项目(E0710018)
