Fault Diagnosis of Chemical Process based on Multivariate PCC Optimization

Abstract:Variables in chemical process are mutually affected and chemical process failures are often caused by the chain effect of a number of variables,so some minor changes in variables can often lead to an unknown fault.In this paper,the multivariable correlations are studied from the perspective of the whole process,and PCC(Pearson Correlation Coefficient) is used to calculate the correlation coefficient.The preliminary optimization on the selected variables are firstly performed.Then,some weighted special variables are selected through mining the effective feature information from the multi-correlation coefficient.Finally,a special variable with large weight is selected from multilayer correlation coefficient set to represent the state of the whole system.In the process of TE(Tennessee Eastman) application,the aggregation weight coefficient Q is used to reduce the fault-tolerance rate of fault diagnosis greatly.The case study shows good effect of correlation coefficient set detection and proves the validity of the method.

Translation or Not:no