CN-122016317-A - Rolling bearing fault identification method under time-varying rotating speed working condition and terminal equipment

CN122016317ACN 122016317 ACN122016317 ACN 122016317ACN-122016317-A

Abstract

The invention belongs to the technical field of fault identification, and particularly discloses a method and terminal equipment for identifying a rolling bearing fault under a time-varying rotating speed working condition, wherein the identification method comprises the steps of acquiring a variable speed vibration signal of the rolling bearing, and determining a plurality of dominant modal components of the variable speed vibration signal; the method comprises the steps of superposing a plurality of dominant modal components, carrying out hierarchical decomposition on the superposed signals to obtain time sequences corresponding to each node of each layer of the superposed signals, determining entropy values of each time sequence by using weighted slope entropy, and inputting the entropy values of each time sequence into a fault recognition model to obtain a fault recognition result. According to the invention, a hierarchical decomposition and weighting strategy is introduced on the basis of slope entropy, so that the amplitude variation of dominant modal components can be quantized, the low-frequency trend component and high-frequency detail information of the signal can be considered, the accuracy and the recognition efficiency of fault feature recognition in the variable-speed vibration signal of the rolling bearing can be remarkably improved, and the sudden shutdown of equipment caused by faults can be avoided.

Inventors

LI ZHE
ZHANG MINHUA
ZHANG CHAO
BAI CHUANXIN
WANG YU
LI LINLIN
WANG CHUNLING
LI YONGJIE

Assignees

西安航空职业技术学院

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (8)

1. The method for identifying the faults of the rolling bearing under the time-varying rotating speed working condition is characterized by comprising the following steps of: step 1, obtaining a variable speed vibration signal of a rolling bearing, and determining a plurality of dominant modal components of the variable speed vibration signal; step 2, superposing a plurality of dominant modal components, and carrying out hierarchical decomposition on the superposed signals to obtain a time sequence of the superposed signals corresponding to each node of each layer; Step 3, determining the entropy value of each time sequence by using the weighted slope entropy; and 4, inputting the entropy value of each time sequence into a fault recognition model to obtain a fault recognition result.
2. The method for identifying faults of rolling bearings under time-varying rotating speed working conditions according to claim 1, wherein the step 1is specifically: Step 1.1, acquiring a variable speed vibration signal of a rolling bearing, and determining a plurality of modal components of the variable speed vibration signal by utilizing variation modal decomposition; and step 1.2, performing modal decomposition on each modal component by utilizing adaptive chirp modal decomposition, and determining a plurality of dominant modal components of the variable speed vibration signal.
3. The method for identifying the rolling bearing faults under the time-varying rotating speed working condition according to claim 2, wherein the method for determining the parameter mode quantity K and the penalty factor alpha in the variation mode decomposition is as follows: And determining the optimal value of the parameter mode quantity K and the optimal value of the penalty factor alpha by using a particle swarm algorithm and taking the minimum envelope entropy of the mode components after the variation mode decomposition as an fitness function.
4. The method for identifying a rolling bearing fault under a time-varying rotating speed working condition according to claim 1, wherein the time sequence of the superimposed signals corresponding to each node of each layer is determined according to a first formula, and the first formula is: ; In the formula, A time sequence of a kth layer p-th node; Is a superimposed signal; For the p-th node corresponding vector in the k-th layer Is an operator of (2); For the p-th node corresponding vector in the k-1 layer Is an operator of (2); for the p-th node corresponding vector in layer 1 Is a part of the operator of (a).
5. The method for identifying a rolling bearing fault under a time-varying rotating speed working condition according to claim 4, wherein the relation between the p-th node and the vector [ ζ 1 , ζ 2 ,…, ζ k ] e {0, 1} is determined according to a second formula, and the second formula is: 。
6. the method for identifying faults of rolling bearings under time-varying rotating speed working conditions according to claim 1, wherein the step 4 is specifically: and inputting the entropy value of each time sequence into the kernel extreme learning machine model to obtain a fault identification result.
7. The method for identifying a rolling bearing fault under a time-varying rotating speed working condition according to claim 6, wherein the method for determining the regularization parameter c and the nuclear parameter g in the nuclear extreme learning machine model is as follows: and determining an optimal value of the regularization parameter c and an optimal value of the kernel parameter g in the kernel extreme learning machine model by using a particle swarm algorithm and taking the classification accuracy larger than a preset threshold as an fitness function.
8. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method for identifying a rolling bearing failure under a time-varying rotational speed condition according to any one of claims 1-7.

Description

Rolling bearing fault identification method under time-varying rotating speed working condition and terminal equipment Technical Field The invention belongs to the technical field of fault identification, and particularly discloses a rolling bearing fault identification method and terminal equipment under a time-varying rotating speed working condition. Background In practical industrial applications, rotating machines are often operated for a long period of time under variable operating conditions, most typically in the form of time-varying rotational speeds. The vibration frequency and the vibration amplitude under the variable working condition can be changed along with the change of the rotating speed, and often with obvious modulation effect, additional frequency components can be introduced, so that the vibration response of the rolling bearing presents stronger non-stationarity and non-linear characteristics, and difficulties are brought to fault characteristic characterization and identification. The self-adaptive decomposition technology based on nonlinear and non-stationary signals can effectively perform self-adaptive decomposition on complex vibration signals. Empirical mode decomposition (EMPIRICAL MODE DECOMPOSITION, EMD) is typically susceptible to mode aliasing, boundary effects, and noise sensitivity. The variational modal decomposition (Variational Mode Decomposition, VMD) can effectively inhibit modal aliasing and end-point effects by establishing a corresponding variational model, but the decomposition performance of the variational modal decomposition is influenced by the number K of parameter modes and a penalty factor alpha, and the parameters often lack prior certainty. Therefore, it is of great importance to develop joint optimization of K and α to obtain a better combination of parameters. In recent years, the nonlinear frequency modulation modal decomposition method converts broadband nonlinear frequency modulation signals into narrowband components equivalently through a demodulation strategy, and realizes multi-component extraction by optimizing and solving. On the basis, related researches are improved aiming at key problems of insufficient stability, bandwidth coefficient updating and the like, and self-adaptive chirp mode decomposition (ADAPTIVE CHIRP Mode Decomposition, ACMD) is provided, so that a more adaptive decomposition framework is provided for nonlinear frequency modulation signal processing. When the rolling bearing fails, its dynamics will change. The entropy analysis can describe the change of the system dynamics complexity and uncertainty from the complex time sequence, and the state monitoring and fault identification of the bearing are realized. The permutation entropy is widely used for vibration feature extraction because of simple calculation and better robustness, but does not consider amplitude information among signals. Slope entropy (Slope Entropy, slE) is a framework based on permutation entropy that quantifies the non-linear characteristics of a time series taking into account the slope between vibration data points and the probability distribution of symbol patterns. However, in the single-scale entropy analysis, key information such as dynamic characteristics of time series at different sizes is easily lost. For this purpose, multi-scale entropy and its modified forms are proposed successively, and the multi-scale method improves the stability of feature extraction by optimizing the conventional coarse-granulation process. However, the multi-scale analysis methods are all based on mean value calculation, and the influence of high-frequency components in the signals is ignored, so that the characteristics of the extracted vibration signals are insufficient. Disclosure of Invention The invention aims to provide a rolling bearing fault identification method and terminal equipment under a time-varying rotating speed working condition, so as to solve the technical problem that the fault identification accuracy is low due to insufficient vibration signal feature extraction in the conventional rolling bearing fault identification method. The first aspect of the invention provides a method for identifying faults of a rolling bearing under a time-varying rotating speed working condition, which comprises the following steps: step 1, obtaining a variable speed vibration signal of a rolling bearing, and determining a plurality of dominant modal components of the variable speed vibration signal; step 2, superposing a plurality of dominant modal components, and carrying out hierarchical decomposition on the superposed signals to obtain a time sequence of the superposed signals corresponding to each node of each layer; Step 3, determining the entropy value of each time sequence by using the weighted slope entropy; and 4, inputting the entropy value of each time sequence into a fault recognition model to obtain a fault recognition result. Preferably, the step 1 spec