On-line large slewing bearing residual life prediction method based on multi-dimensional data drive

Abstract

The invention discloses an on-line large slewing bearing residual life prediction method based on multi-dimensional data drive. The method includes the steps of firstly, conducting a full-life fatigue test on a slewing bearing, and obtaining multiple sets of vibration acceleration signal monitoring data of the whole test cycle of the slewing bearing; secondly, decomposing a vibration signal into a plurality of intrinsic mode functions (IMFs) through ensemble empirical mode decomposition (EEMD), and selecting a plurality of IMFs which can reflect performance degradation of the slewing bearing for signal reconstruction through principle component analysis (PCA) so that the aim of filtering noise elimination can be achieved; thirdly, conducting PCA on multi-dimensional acceleration signals at different stages and signals of the initial period of the test so as to obtain one-dimensional continuous SPE (C-SPE), reflecting the change situations of the multi-dimensional acceleration vibration signals on the basis of the one-dimensional C-SPE, calculating multiple time domain characteristics of the C-SPE, and establishing a performance degradation model of the slewing bearing. The method is few in manual intervention process, and the prediction result is closer to the engineering practice.

Description

technical field [0001] The invention belongs to the fields of vibration signal processing and equipment health monitoring, and relates to an online prediction method for the remaining service life of a large slewing bearing driven by multidimensional data. Background technique [0002] Large-scale slewing bearings are often used as the core slewing connection in large and heavy equipment such as wind turbines, excavators, and cranes. Their serious failures will cause huge losses. Therefore, realizing online remaining life prediction of slewing bearings is of great significance for formulating effective active maintenance or replacement strategies and improving production efficiency. [0003] Large-scale slewing bearings generally work under the conditions of large background noise, low speed and heavy load. The fault passing frequency in the non-stationary random signal generated by each component can be as low as 1Hz and the energy is very low, which leads to the effective ...

AI Technical Summary

Problems solved by technology

[0005] However, EEMD is used in common slewing bearing vibration signal fault processing methods only to decompose the signal, and then analyze the energy of each IMF, and seldom filter and select the IMF to reconstruct the signal. The EEMD-MSPCA proposed by foreign scholars can effectively Perform filtering and denoising, but its effect de

Images