Traditional multivariate statistical fault detection methods are designed for single operating conditions and may produce erroneous conclusions if used for multi-mode process monitoring.A novel multi-mode process monitoring approach based on PCA mixture model is proposed in this paper.First

the PCA technique is used to reduce the dimension of original variables and to guarantee the nonsingular covariance matrix.In order to overcome the limitations of EM (expectation-maximization)

the BYY(Bayes Ying-Yang)scale-incremental EM algorithm is then adopted to automatically optimize the number of mixture components.With the obtained PCA mixture model

a novel process monitoring scheme is derived for fault detection of multi-mode processes.The validity and effectiveness of the proposed monitoring approach are illustrated by numerical example and the TE process.The results show that the proposed algorithm can achieve good parameter estimation of the mixture model with correct model selection.Therefore

it can achieve accurate and early detection of various types of faults in multi-mode processes.