基于多项式滤波的化工过程数据预处理  

Error Reduction of Chemical Process Data Based on Polynomial Filters

作　　者：黄兆杰[1] 金思毅[1] 王莉[1] 

机构地区：[1]青岛科技大学化工学院,山东青岛266042

出　　处：《青岛科技大学学报（自然科学版）》2012年第3期269-272,276,共5页Journal of Qingdao University of Science and Technology:Natural Science Edition

摘　　要：用多项式滤波预处理化工过程数据时,多项式滤波的除噪效果与多项式次数和数据点数目有关。当选取的数据点数目不大时,提出一种根据中位数选取多项式次数的方法,以提高多项式滤波的除噪效果。基于该方法,同时研究了多项式滤波的过失误差侦破性能。选取一反应釜系统进行仿真运算,计算结果表明,基于中位数选取多项式次数的方法能有效地除噪、有效地侦破过失误差。When polynomial filters are used for error reduction of chemical process data, the error reduction is influenced by the number of data points used and the order of poly- nomial filters. When the number of data points is not large, a method of choosing the order of polynomial filters is proposed to achieve a good error reduction in this investiga- tion. Based on this method, the performances of gross errors detection of polynomial fil- ters are used in a agitated reactor system. Simulation results show that this new method is characterized by error reduction and gross errors detection.

关 键 词：多项式滤波 多项式次数选取 除噪 过失误差侦破 

分 类 号：TQ015.9[化学工程] TP274[自动化与计算机技术—检测技术与自动化装置]