改进多维本质时间尺度分解的厂级振荡检测  

Plant-wide Oscillation Detection Based on Improved Multivariate Intrinsic Time-scale Decomposition

作　　者：王宇红[1] 高志兴 WANG Yu-hong;GAO Zhi-xing(School of Control Science and Engineering,China University of Petroleum(East China),Qingdao 266580,China)

机构地区：[1]中国石油大学(华东)控制科学与工程学院,山东青岛266580

出　　处：《控制工程》2022年第10期1835-1840,共6页Control Engineering of China

摘　　要：随着现代工业的发展,控制回路的厂级振荡检测已成为控制性能评估的重要研究内容。针对多维本质时间尺度分解(MITD)算法在分解复杂的多维信号时无法较好地提取原信号的局部特征的问题,提出了一种改进的多维本质时间尺度分解(IMITD)。该算法通过对信号进行端点延拓、选取多维信号基线节点和增加筛选内循环的方法,更好地保留了原信号局部特征和时频信息,在控制回路的厂级振荡检测上提高了检测的准确性。仿真信号和工业实例验证了改进的多维本质时间尺度分解算法在工业多回路振荡检测中的优越性。With the development of modern industry, the automatic detection of plant-wide oscillations in control loops is becoming one of the important topics in control performance evaluation. Aiming at the problem that the multivariate intrinsic time-scale decomposition(MITD) algorithm cannot extract the local features of the original signal well when decomposing complex multidimensional signals, an improved multivariate intrinsic time-scale decomposition(IMITD) algorithm is proposed in this paper. In this work, improvements are made in the following three aspects, endpoint extension of the original signal, selection of multivariate baseline-nodes and increase of filtering inner loops. Compared with the multivariate intrinsic time-scale decomposition algorithm, the improved multivariate intrinsic time-scale decomposition algorithm can better retain the local characteristics and time-frequency information of the original signal, and improves the detection accuracy of plant-wide oscillations in control loops. The advantages of the improved multivariate intrinsic time-scale decomposition algorithm in industrial plant-wide oscillation detection are proved by simulation signals and industrial examples.

关 键 词：多维信号处理 改进多维本质时间尺度分解 厂级振荡检测 模态混叠 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]