Chemical process failure detection method based on failure-dependent principal component space

一种基于故障相关主成分空间的化工过程故障检测方法

Abstract

本发明公开了一种基于故障相关主成分空间的化工过程故障检测方法,所述方法包括:根据正常状态下的历史数据采用PCA构建主元空间和残差空间;根据故障状态下的历史数据采用GA对主元空间进行优化得到故障相关主元;在各个故障相关主元子空间及残差空间内构造统计量;采用贝叶斯方法将各个故障相关主元子空间及残差空间的统计量融合成综合统计量;在线监测时,根据在线采集数据计算综合统计量,判断运行状态。本发明利用历史正常数据建立故障检测空间,利用历史故障数据采用GA对故障检测空间进行优化,降低故障检测的冗余,提高故障检测的效率与准确率。
The invention discloses a chemical process failure detection method based on a failure-dependent principal component space. The method comprises the following steps: constructing a principal component space and a residual space by PCA (Principal Component Analysis) according to history data in a normal state; optimizing the principal component space by a GA (Genetic Algorithm) according to history data in a failure state to obtain a failure-dependent principal component; constructing statistics in each failure-dependent principal component sub-space and the residual space; fusing the statistics of each failure-dependent principal component sub-space and the residual space into comprehensive statistics by a Bayes method; and calculating the comprehensive statistics according to online acquired data during online monitoring, and judging a running state. Through adoption of the chemical process failure detection method, a failure detection space is constructed by history normal data, and the failure detection space is optimized by the GA with history failure data, so that redundancy of failure detection is lowered, and the efficiency and accuracy of the failure detection are increased.

Claims

Description

Topics

Download Full PDF Version (Non-Commercial Use)

Patent Citations (0)

    Publication numberPublication dateAssigneeTitle

NO-Patent Citations (0)

    Title

Cited By (0)

    Publication numberPublication dateAssigneeTitle