A Fault Monitoring Method Based on Hierarchical Density Peak Clustering and Most Similar Modes

Abstract

The invention relates to a fault monitoring method based on hierarchical density peak clustering and the most similar mode. The historical normal data of an industrial process are modally divided to acquire hierarchical modal information. The hierarchical modal information is used to establish a fault monitoring model for the historical normal data of the industrial process. The most similar modeof the industrial process data to be monitored is acquired and input into the fault monitoring model for fault monitoring. According to the invention, existing industrial data resources are used; multimodality and multimodal dynamics and uncertainty of a complex industrial process are considered; the limitations of relying on priori modal information and using fixed modal dividing and models of the existing multimodal fault monitoring method are overcome; and the method is important for timely detecting abnormal conditions in the industrial process, ensuring production safety and improving product quality.

Description

technical field [0001] The invention relates to the technical field of fault monitoring and diagnosis, in particular to a fault monitoring method based on hierarchical density peak clustering and most similar modes. Background technique [0002] The purpose of fault monitoring is to discover abnormal working conditions of industrial processes in time, ensure production safety, and improve product quality. The operating conditions of industrial processes are usually changed due to changes in raw materials, manufacturing parameters, product specifications, etc., resulting in multiple operating modes. The different modes have similar, each specific characteristics and durations. A mode is defined as a process with similar statistical properties and a certain duration. In addition, there may be a gradual transition process between two adjacent modes. Therefore, it is of great significance to study effective and feasible multi-mode process fault monitoring methods. [0003] M...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention provides a fault monitoring method based on hierarchical density peak clustering and most similar modes, which considers the multi-modality and multi-modal dynamics and uncertainty of complex industrial processes, and overcomes the Problems of Existing Multimodal Fault Monitoring Methods Relying on Prior Modal Information

Images