Fault prediction and anomaly diagnosis of permanent magnet fast ring cabinet based on Bayesian network inference modeling

In the field of industrial process control in recent years, fault diagnosis technology has just become a very important and popular research direction.By installing sensors in the permanent magnet fast ring cabinet to collect multi-source data, and using data cleaning and standardization techniques, a fault detection and diagnosis method combining Bayesian inference MCMC and DPMM is proposed.The Bayesian inference MCMC method is the core of this fault correction technique, which is analyzed by comparing the differences between the observed data and various types of fault data.The results show that the CDC and RBC methods do not have the ability to track the fault propagation process, while the method proposed in this paper is able to analyze the different fault variables in the propagation process of faults, which verifies the robustness and practicability of this paper's method from multiple perspectives.