Failure Prediction For Industrial Process Based On Correlation Analysis Of Operating Parameters

Designed to ensure the safety and reliability of industrial equipment during operation and reduce maintenance costs,Prognostics and Health Management(PHM)has received widespread attention over the last decades and has had a profound impact on industry.With the increasing complexity of industrial equipment and the continuous development of sensor technologies,fault prediction based on multi-sensor information has become the frontier technology.During the operation of industrial equipment,a large number of operating parameters reflect the operating state of the equipment simultaneously,and these operating parameters are correlated with each other.Therefore,we will analyze the correlation of operating parameters and predict the failure of industrial equipment based on the analysis results.First,we review the related work about fault prediction technologies with special introduction to the fault prediction method based on multi-sensor information,and analyze the advantages and disadvantages.Then,three different correlation analysis methods are proposed,i.e.,correlation analysis based on distance function,correlation analysis based on association rule mining and correlation analysis based on dynamic association rule mining.Based on the results from correlation analysis,we provide failure prediction for industrial equipment with the autoregressive model or artificial neural networks.The proposed methods are verified using some industrial cases,and the experimental results validate the effectiveness of the proposed methods.Finally,we summarize the thesis and briefly introduce the future research.