Fault detection CVA algorithm of chemical process based on ISOMAP

Strong auto-correlation and cross-correlation in chemical processes data can be dealt well by canonical variate analysis (CVA) algorithm, but this algorithm can´t solve nonlinear problem of chemical process data. So a fault diagnosis CVA algorithm of chemical process based on isometric feature mapping (ISOMAP) is proposed in this paper. At first, this algorithm uses ISOMAP algorithm of manifold learning to achieve realize nonlinear dimensionality reduction for initial data and maintain internal geometry structure of data. Then CVA is used to the extracted low dimensional data to obtain process state space description and SPE statistics. Fault detection simulation results of TE process show that the proposed algorithm is more effective to detect faults of chemical process than CVA algorithm.