A Fault Diagnosis Method for Offshore Platform Air Compressor Based on LSTM

Abstract

The invention belongs to the technical field of fault diagnosis of marine engineering equipment, in particular to an LSTM-based method for fault diagnosis of an air compressor on an offshore platform. The present invention constructs the multi-element sensor signals of the offshore platform air compressor in a time series sequence, provides efficient and reasonable data structure support for Kalman filtering and later LSTM model construction, is conducive to improving the denoising effect of the previous linear filtering, and has It helps to extract and retain reasonable fault features in the LSTM weight matrix later. Through the combination of digital signal processing and deep learning theory in the field of artificial intelligence, the present invention transforms the complex fault diagnosis process of the air compressor on the offshore platform into an easy-to-use diagnosis method based on multi-sensor monitoring, and through the fusion innovation of the algorithm level, the maximum The hardware requirement of the diagnosis process is reduced to a certain extent, the diagnosis time and diagnosis cost are reduced, and more precious time is gained for fault repair.

Description

technical field [0001] The invention belongs to the technical field of fault diagnosis of marine engineering equipment, in particular to an LSTM-based method for fault diagnosis of an air compressor on an offshore platform. Background technique [0002] With the rapid development of the world economy, the demand for energy has increased sharply, and the scale of oil and gas extraction has been expanding day by day. From the 1880s to the present, technological progress and breakthroughs have accelerated the progress of offshore oil and gas exploitation. During this process, all kinds of production equipment on the offshore platform show a trend of large-scale, highly correlated, complex and automated. In the actual working process, the system equipment of the offshore platform will produce various types of failures. Under the influence of the "domino effect", other related systems of the platform will also experience unpredictable failures, which will cause major economic an...

AI Technical Summary

Problems solved by technology

However, the above methods have certain shortcomings, and cannot well solve the problems of complex fault signal excitation and strong nonlinear fault mapping in the fault diagnosis of platform air compressors, which greatly limits the fidelity and accuracy of fault signal acquisition. The accuracy of fault diagnosis increases the misdiagnosis rate and brings many uncertainties to equipment maintenance and health management

Images