Monitoring Method of CIPCA Grinding Process Based on Multi-modal Data Characteristics

In order to solve the problems of multi-work and uncertainty in the grinding process, a process monitoring method called Multimodal Complete Information Principal Component Analysis was proposed in this paper. In this method, SDAE-K-Means++ is adopted to carry out modal partitioning of grinding production process data, and online mode recognition is realized by calculating the similarity between mode center and online data. At the same time, considering the uncertainty characteristics of field data due to noise and sensor drift, CIPCA was combined with modal center estimation for monitoring. This method combined interval data processing with interval monitoring method to reduce the impact of uncertain information. By applying to the actual grinding process of a mine to verify the effectiveness of the proposed method.