A monitoring method based on combination of EPCA and RPCA

Process monitoring is a very important measure which ensures process safety and stable operation. Considering the insufficient process information at the very beginning of new process and the necessity that process model needs to be updated constantly in order to adapt to process changes, a method based on a combination of extended principal component analysis (EPCA) and recursive principal component analysis (RPCA) is proposed in this article. Afterwards, the method proposed is applied in grinding processes and simulated, which shows that the method can not only effectively solve the problem of critical shortage of process information when a new process is run, but also improve the utilization rate of the normal followed-up samples, making process model more adaptable to process variability. Finally, the simulations results illustrate effectiveness and practicality of the investigated method.