包含阀门非线性的闭环系统性能评价  

Performance assessment of closed-loop system with valve nonlinearity

作　　者：李涛 黄春庆 LI Tao;HUANG Chunqing(School of Aerospace Engineering,Xiamen University,Xiamen 361102,China)

机构地区：[1]厦门大学航空航天学院,福建厦门361102

出　　处：《厦门大学学报（自然科学版）》2020年第4期522-527,共6页Journal of Xiamen University：Natural Science

基　　金：深圳市科技计划项目(JCYJ20180306172924636)。

摘　　要：当闭环系统采用PI或PID控制器时,控制阀黏滞的存在容易引起振荡.将控制性能评价(CPA)推广到这类控制系统中并非易事.然而,就像线性系统一样,最小方差(MV)反馈不变量的存在使得CPA在这类系统中成为可能.在此基础上,该文采用经验模态分解(EMD)方法将闭环系统发生振荡时的线性随机分量从输出数据中分离出来,利用改进的滤波和相关性分析(FCOR)算法估计出CPA的MV基准.数值仿真实验证明了该方法的有效性.When PI or PID controller is employed,the control-valve stiction often causes oscillations.It is difficult to extend control performance evaluation(CPA)to such control systems.Fortunately,just like linear systems,the existence of minimum variance feedback invariants makes CPA possible in such systems.In this paper,minimum variance benchmark of CPA is estimated by using the proposed modified filtering aud correlation analysis(FCOR)algorithm after the empirical mode decomposition(EMD)method is employed.Then the linear stochastic component is separated from the output data of the closed-loop system when oscillations occur.Meanwhile,the valve-stiction characteristic is identified as a linear model and accordingly the benchmark is corrected further.Numerical simulations illustrate the effectiveness of the proposed method.

关 键 词：控制性能评价 最小方差控制 阀门黏滞 经验模态分解 滤波 相关性分析 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]