支持向量机在PID控制器参数整定中的仿真研究

Simulation on Tuning of PID Parameters Based on Support Vector Machine

PID控制是应用最为广泛的控制方法,由于系统中存在非线性和时变性,影响建立精确的模型,系统性能。为了解决控制参数整定,改善系统性能,提出一种基于支持向量机的PID控制器参数整定方法。通过将支持向量机和PID控制器相结合建立支持向量机的参数整定模型,在控制过程中将PID控制的参数作为支持向量机的输入,构造参数自适应学习的PID控制器,在控制过程中动态调整PID的三个控制参数,进行仿真的在线整定。仿真结果表明,支持向量机的PID控制方法在处理非线性和时变系统时,提高了实时性能,增强系统稳定性,并获得更好的控制效果,为通用非线性PID控制器设计提供了依据。

PID controllers are most widely used in process control industries.But in actual application parameter setting is still not get better solution.In order to solve the control parameters amd improve system performance,a method is proposed based on support vector machine（SVM） for PID controller,and parameter setting model is bulit combined support vector machine（SVM） and PID controller.In the control process of PID,control parameters is used for support vector machine input,and parameters can be adjusted in control process and thus sets PID parameters on line.Simulation results indicate that the PID controller based on SVM has stronger adaptability and better results.The presented control strategy introduces a new way for nonlinear control.