Visual feature fusion-based complex rotating mechanical equipment fault diagnosis method

Abstract

The invention discloses a complex rotating mechanical equipment fault diagnosis method based on visual feature fusion and relates to the technical field of rotating mechanical equipment fault diagnos. The method comprises steps that firstly, a virtual vibration signal is obtained through an original vibration signal, and a vibration image representing an aero-engine fault is constructed through the virtual vibration signal and the original vibration signal; high-level fault features are obtained by adopting deep learning and transfer learning, finally, in a continuous sampling period, multi-modal fault features are clustered to form a fault feature group, a multi-modal fault feature space-time model is constructed based on space-time correlation, and an effective tensor decomposition method is proposed to realize fault diagnosis under the model. The complex rotating mechanical equipment fault diagnosis method based on visual feature fusion solves problems that an existing diagnosis method is incomplete in fault diagnosis sample data, difficult in fault feature representation extraction and poor in fault diagnosis capability.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis of rotating mechanical equipment, in particular to a fault diagnosis method for complex rotating mechanical equipment based on fusion of visual features. Background technique [0002] Aeroengine is a kind of highly complex and precise thermal machinery. As the heart of aircraft, it is not only the power of aircraft flight, but also an important driving force to promote the development of aviation industry. Every important change in the history of human aviation is closely related to the technological progress of aeroengine Inseparable. [0003] Fault diagnosis technology is a powerful guarantee for the safe and efficient operation of complex rotating machinery, but the implementation of this type of technology still faces many challenges: due to the small number of vibration signal samples collected by the aeroengine vibration sensor, noise interference and non-stationarity, etc. The perf...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a complex rotating mechanical equipment fault diagnosis method based on visual feature fusion to solve the problem of incomplete fault diagnosis sample data, difficulty in extracting fault feature representations and fault diagnosis capabilities in the existing diagnostic methods proposed in the background technology poor problem

Images