A Fault Diagnosis Method Based on Adaptive Manifold Embedding and Dynamic Distribution Alignment

Abstract

The invention discloses a fault diagnosis method based on adaptive manifold embedding and dynamic distribution alignment. The method automatically calculates the optimal subspace dimension, calculates the geodesic flow kernel and the transformed manifold feature representation, It can effectively avoid the characteristic distortion of the data in the original Euclidean space. Introduce the similarity measure A‑distance to define an adaptive factor, dynamically adjust the relative weight of the conditional distribution and edge distribution of the sample data, effectively narrow the distribution difference between the source domain and the target domain samples, and greatly improve the accuracy of rolling bearing faults under variable working conditions. The accuracy and validity of the diagnosis, the method has strong interpretability, lower requirements on computer hardware resources, faster execution speed, and excellent diagnostic accuracy, algorithm convergence and parameter robustness. This method is especially suitable for multi-scenario and multi-fault bearing fault diagnosis under variable working conditions, and can be widely used in fault diagnosis tasks under variable working conditions of complex systems such as machinery, electric power, chemical industry, and aviation.

Description

technical field [0001] The invention relates to the technical fields of mechanical fault diagnosis and machine learning, in particular to a fault diagnosis method based on adaptive manifold embedding and dynamic distribution alignment. Background technique [0002] Rotating machinery is ubiquitous in modern industrial production. As a key component in industrial production equipment, rolling bearings are widely used in various important fields such as machinery, electric power, chemical industry, aviation, etc. At the same time, rolling bearings often work under high temperature, high speed, heavy load In such harsh environments, failures such as wear, cracks, and fractures are prone to occur. Once the bearing fails, it will bring huge economic losses, and cause catastrophic accidents such as casualties. Therefore, it is necessary to strengthen the condition monitoring capabilities of mechanical equipment and improve It is of positive and great significance to qualitatively ...

AI Technical Summary

Problems solved by technology

Furthermore, when adaptive distribution alignment is performed in the original Euclidean space, data feature distortions inevitably occur, leading to suboptimal diagnostic performance of the model

Images