Research On Fault Diagnosis Of Hot Strip Mill Process Based On Multivariate Statistical Method

The fault of hot strip mill process is related to quality and safety.Due to the complexity of hot mill process,it is impossible to establish accurate model,so the fault diagnosis method based on multivariate statistics is adopted.This method does not need to establish the model accurately,but only needs to analyze the data to judge the system.In view of the characteristics of non-linear,strong correlation and non Gaussian distribution of hot strip mill process data,a canonical independent component analysis method is proposed.This method combines the advantages of canonical variable analysis to remove the correlation,independent component analysis can extract independent components from non Gaussian distribution data,and is suitable for nonlinear systems.The results show that the alarm rate of fault detection is 100%,and the average false alarm rate is reduced by 22.4%.In view of the time-varying process data of hot strip mill process,a canonical variable analysis method based on moving window is proposed.This method can update the monitoring statistics and control limits in real time,so as to improve the accuracy of fault detection.The fault detection alarm rate is 100%,and the average false alarm rate is reduced by 8.88%.According to the characteristics of non-linear,strong correlation,time-varying and non Gaussian distribution of hot strip mill process data,the moving window canonical independent component analysis method is proposed.The method combines the moving window with the canonical independent component analysis method to further improve the accuracy of fault detection.According to the results,the false alarm rate is only 0.2%.In order to solve the problem that the fault cause can not be found out in time,the above improved method is combined with the contribution plot.The results show that moving window canonical independent component analysis method can find out the cause of the fault,which is consistent with the actual situation.Figure 29;Table 4;Reference 60...