A Fault Diagnosis Method for Multimodal Process Quality Correlation Based on Sparse GMM

Abstract

The invention discloses a multi-mode process quality-related fault diagnosis method based on sparse GMM, which uses sparse representation to obtain high-quality coefficient weight matrix, and integrates manifold structure information to construct a sparse Gaussian mixture model, so that the probability distribution of Gaussian components Smooth changes along the data manifold structure, and similarity between local neighbor samples of Gaussian components, automatically obtain the number of Gaussian components, robust to noise and outliers, obtain quality-dependent fault detection, and at the same time rely on the detected The controlled neighbor of the fault locates the root variable of the fault. Compared with the Gaussian mixture model monitoring method, the method of the invention characterizes the local manifold structure of the process data and the sparse relationship of the data, obtains the local similarity relationship between samples, and reflects the change of the multi-modal process. Therefore, the sparse GMM method involved in the present invention can achieve better fault detection effect and accurately locate the root variable of fault occurrence.

Description

technical field [0001] The invention relates to the technical field of industrial process monitoring, in particular to a multi-mode process quality-related fault diagnosis method based on sparse GMM. Background technique [0002] Process monitoring in modern industry plays a pivotal role in ensuring production safety and increasing production. With the development of distributed control systems, the scale of production and the complexity of operations have increased dramatically, and a large amount of high-dimensional data has been collected during the process. Moreover, because the product grade and output will be constantly adjusted with market demand and seasonal effects, and process parameters such as product composition, process settings, and feed ratios will also fluctuate, modern industrial processes will operate in multiple different operating modes. switch between states. These random changes in the production process make the process data present characteristics ...

AI Technical Summary

Problems solved by technology

Moreover, it is difficult to directly measure the key variables such as output and product quality in the production process online, but after the production is completed

Images