交流电力测功机的支持向量机模糊PID控制策略

SVR-fuzzy-PID strategy of AC electrical dynamometer

为了提高交流电力测功机运行的稳定性和控制精度,将支持向量回归预测(support vector regression,SVR)、模糊控制、PID控制等相结合,提出一种支持向量机模糊PID控制策略(SVR-fuzzy-PID)。该控制策略首先建立模糊规则,并通过模糊规则调节PID控制器的比例系数、积分系数、微分系数,然后,通过PID控制器控制输出量,在反馈环节利用支持向量回归预测模型预测控制对象在一定时间后的反馈量,并将预测值代替反馈值,从而实现“超前”控制,以解决传统控制系统的控制滞后问题。研究结果表明:该控制策略设计的交流电力测功机系统控制系统比传统的控制系统具有更加优良的性能,可实现交流电力测功机系统的快速和稳定控制。

In order to improve the stability and control accuracy of alternating current(AC)electrical dynamometer,a SVR-fuzzy-PID based control strategy was proposed through combing the theories of SVR(support vector regression)prediction,fuzzy control and PID control.In this control strategy,fuzzy rule was firstly established,which was used to adjust scale,integral and differential coefficients to control the output of PID controller.The SVM(support vector machine)prediction model was applied to predict the feedback of controlled objects to realize the advanced control and solve the problem of control delay.The results show the proposed control strategy presents a better control performance for alternating current electrical dynamometer compared to traditional control strategy,which can realize the fast and stable control of alternating current electrical dynamometer.