Construction of spectral clustering adjacency matrix based on L3CRSC and application thereof

Abstract

The invention discloses a construction method of a spectral clustering adjacency matrix based on L3CRSC and the application thereof, the adjacency matrix is constructed by employing a latent-space low-rank representation coefficient, and the coefficient is input to spectral clustering to realize separation and recognition of different kinds of data features. According to the method which is different from a conventional LRR algorithm, a clean dictionary matrix is obtained through separation of original data, and the low-rank coefficient is learned by employing the dictionary; and an effective dictionary and an effective matrix in the latent space can be obtained through update by employing the latent-space low-rank representation algorithm. A new method for constructing a latent-space mapping matrix is also provided, in which the local geometric structures of retained data before and after mapping are consistent. According to the method, during recognition of real bearing fault data, different types of fault features can be effectively separated, the recognition accuracy is improved, and the guarantee is provided for later fault diagnosis.

Description

technical field [0001] The invention relates to the construction and application of spectral clustering adjacency matrix based on L3CRSC. Background technique [0002] 1. Unsupervised data identification based on spectral clustering is a research hotspot in recent years, and its key task is to construct an adjacency matrix. In this paper, we propose a new unsupervised identification algorithm based on spectral clustering, called Locally Constrained Latent Space Low Rank Subspace Clustering (L3CRSC). The adjacency matrix is ​​constructed by using the low-rank representation coefficients of latent space, and input into spectral clustering to realize the separation and identification of different types of data features. Different from the traditional LRR algorithm, we obtain a clean dictionary matrix by separating the original data, and use the dictionary to learn low-rank coefficients. In order to better deal with high-dimensional features, we use the latent space low-rank r...

AI Technical Summary

Problems solved by technology

[0003] 2. The identification of fault data is an important task of bearing fault diagnosis. When two or more faults occur at the same time during the operation of the bearing, these faults are often coupled with each other, which increases the difficulty of fault identification

[0008] 7. Fault data are often distributed in a series of unrelated or overlapping high-dimensional subspaces. Directly using the eigenvalues ​​to calculate the corresponding representation coefficients and construct the adjacency matrix will cause the degradation of the diagonal block structure

However, the algorithm uses the PCA method to directly project the original data into the latent space, which is not effective for some nonlinear data.

Images