A Pipeline Status Recognition Method Based on Multi-attribute Feature Fusion of Monitoring Data

Abstract

The invention discloses a pipeline state recognition method based on the fusion of multi-attribute features of monitoring data. A state recognition framework is constructed based on the pipeline state monitoring data, so that the framework includes normal, blocked and leaking state modes. Based on the state modes contained in the state recognition framework, Extract the data characteristics of the pipeline operation status; use the evaluation index of comprehensive sensitivity and volatility to select the above data characteristics, use the obtained characteristics as the evidence body of the status identification framework, and process the status data of the pipeline to be identified to obtain the pipeline status. Identify the corresponding evidence body of the state data, obtain the basic probability value of the evidence body, use the synthesis rules of the D-S evidence theory, conduct fusion analysis on all the evidence information of the current state, determine that the current state belongs to the type in the identification framework, and realize the operation of the pipeline Status identification and judgment. The invention proposes a pipeline running state identification technology with guiding significance, which is of great significance for improving the accuracy of technicians' decision-making and judgment on the pipeline running state.

Description

technical field [0001] The invention belongs to the technical field of pipeline operation state identification, and in particular relates to a pipeline state identification method based on multi-attribute feature fusion of monitoring data. Background technique [0002] Pipelines are unit parts used to transport liquids, gases, and other substances. They are an important part of modern industrial production systems and are widely used in complex production systems such as petroleum, chemical, and water conservancy. An important part of. Because pipelines usually transport flammable, explosive, and toxic media, and are affected by factors such as load-carrying stress, corrosion of transported media, and aging of parts, they are prone to leakage, blockage, and other failures, which often cause significant personnel and property losses. Therefore, how to use the condition monitoring data during pipeline service to realize the identification and discovery of faults such as small...

AI Technical Summary

Problems solved by technology

However, this method is limited by the accuracy and detail of the operating mechanism of the pipeline system, limited by empirical knowledge and modeling knowledge, and lacks the effective use of a large amount of accumulated pipeline operation status and failure mode data; at the same time, the process of pipeline failure It is a small and gradual process, how to effectively extract the feature information of the failure mode, how to evaluate and select the high-quality features that characterize the pipeline operating state, and conduct information fusion analysis have become problems that need to be solved

Images