Gearbox fault diagnosis method based on multi-element modal decomposition-transfer entropy

Abstract

The invention relates to a gearbox fault diagnosis method based on multi-element modal decomposition-transfer entropy, and belongs to the technical field of gear fault analysis. According to the technical scheme, a multi-element modal decomposition noise reduction algorithm is adopted to extract system main characteristics firstly, then transfer entropy is used for describing system complexity, and is used for gear fault diagnosis. According to the gearbox fault diagnosis method, noise-assisted multi-element empirical modal decomposition-transfer entropy is used to analyze gearbox fault vibration information transmission characteristics, nonlinear coupling and information transmission characteristics between output shaft end and input shaft end signal frequency bands under a gearbox faultcondition are quantitatively described so that a transmission path of gearbox fault vibration signals can be explored, a gearbox state evaluation index based on the multi-element empirical modal decomposition-transfer entropy is established, and a new effective means is provided for fault diagnosis, performance degradation state identification and trend prediction of a rotary machine.

Description

technical field [0001] The invention relates to a gearbox fault diagnosis method based on multivariate mode decomposition-transfer entropy, belonging to the technical field of gear fault analysis. technical background [0002] The study of practical and reliable mechanical equipment fault diagnosis technology is the technical basis for preventing mechanical equipment failure and ensuring the safe and stable operation of mechanical equipment. As a kind of important technical equipment widely used in the field of national defense and national economy, the mechanical power transmission system is of great importance to its safety and reliability. As a key component of the power transmission system, gears work under high load and high speed for a long time, and are vulnerable to damage and failure. Their failure will directly affect the operation of the entire mechanical system. Therefore, it is necessary to improve the reliability of rotating machinery and detect failures in a t...

AI Technical Summary

Problems solved by technology

Although approximate entropy only needs less data to measure the complexity of time series, its calculation will be biased due to the inherent comparison of approximate entropy with its own data segments, which leads to its small complexity to the system. less sensitive to changes

At the same time, these algorithms all require sufficient data length, while the data length of rotating machinery vibration signals is generally limited

Therefore, these nonlinear dynamic indices are inconvenient in practical engineering applications. It is necessary to find a new method that is more effective, stable and requires less data length. The above-mentioned information entropy method cannot comprehensively and accurately describe and quantitatively describe gearbox faults. In this case, the nonlinear coupling and information transmission characteristics between the output shaft end and the input shaft end signal frequency band

Images