模糊神经网络在阀门开度控制中的应用

Application of Fuzzy Neural Network in Valve Opening Control

针对电磁阀控气缸系统时变、非线性的特性,应用模糊神经网络整定PID参数。利用AMESim物理图形建模软件搭建被控对象的模型,应用MATLAB编写模糊神经网络PID的S函数,再联合MATLAB和AMESim对算法进行仿真,仿真结果表明模糊神经网络PID的适应性好、抗干扰能力强,性能优于常规PID。最后搭建实验平台进行了实验。实验结果表明,系统在不同环境下的控制精度高、响应快、过程平稳,满足阀门开度控制的要求。

According to the characteristics of time varying and nonlinear of solenoid valve control cylinder pneumatic system,the parameters of PID were adjusted by fuzzy neural network control.The modeling process of the controlled object was simulated by using the AMESim physical graphics modeling software,the S function of fuzzy neural network PID was established by MATLAB,then MATLAB and AMESim were combined to simulate the algorithm.The simulation results show that fuzzy neural network PID has good adaptability and strong anti-interference,the control performance is better than the routine PID control.Finally,an experimental platform was built,and experiments are carried out.The experimental results show that the system has high control accuracy,fast response and stable process under different environments,which can meet the requirements of valve opening control.