Fault diagnosis based on batch-wise unfolded MPCA of papermaking wastewater treatment process

In this paper, a method based on batch-wise unfolded Multi-way Principal Component Analysis (MPCA) is evaluated for process fault diagnosis analysis of a pilot-scale SBR aerobic stage. This method separates the observation space into a principal component subspace (PCS) and a residual subspace (RS). Firstly, faulty batches are observed if the values of T
²
and SPE exceed the control limits; secondly, the faulty time can be detected by calculating SPE of every sampling point in the faulty batch; lastly, the fault elements which cause the fluctuation can be diagnosed with contribution plots. Research results show that the batch-wise unfolded MPCA can be used to monitor the papermaking wastewater treatment process and identify the process fault efficiently.