A novel strategy of monitoring batch process based on Mean vector component analysis

A novel monitoring strategy based on Multi-way Mean vector component analysis (MMVCA) is proposed for the online fault detection of batch process. The faults that affect quality index are denoted as quality-related faults, which should be taken care of as soon as possible. The method is based on dimensionality reduction by preserving the squared length, and implicitly also the direction, of the mean vector of the original data. The optimal mean vector preserving basis is obtained from the spectral decomposition of the inner-product matrix, and it is shown to capture clustering structure. Unlike traditional Multi-way Principal Component Analysis (MPCA), these axes are in general not corresponding to the top eigenvalues. The proposed algorithm has been applied in penicillin fermentation system and plant data, to verify the effectiveness of the method.