Bearing fault diagnosis method based on improved ant lion algorithm and support vector machine

Abstract

The invention discloses a bearing fault diagnosis method based on an improved ant lion algorithm and a support vector machine. The bearing fault diagnosis method comprises the following steps that vibration acceleration signals under a typical fault state are collected; the collected signals are extracted to obtain a data sample of a typical fault type; based on the improved escape mechanism ant lion optimization algorithm and the support vector machine, a bearing fault diagnosis model is established; the data sample is input into the bearing fault diagnosis model to optimize the bearing faultdiagnosis model; and bearing fault diagnosis is performed based on the optimized bearing fault diagnosis model. According to the bearing fault diagnosis method based on the improved ant lion algorithm and the support vector machine, the improved EALO algorithm is provided by introducing an escape mechanism and adaptive convergence conditions to improve the optimization performance of the algorithm, and the improved ant lion algorithm is combined with the support vector machine to realize the bearing fault diagnosis, so that important theoretical significance and practical value are achieved on improving the rolling bearing fault diagnosis accuracy, ensuring the safety of rolling bearings and stabilizing the operation.

Description

technical field [0001] The invention relates to the field of mechanical fault diagnosis, in particular to a bearing fault diagnosis method based on an improved ant lion algorithm and a support vector machine. Background technique [0002] Rolling bearings are core basic components and are widely used in important mechanical equipment such as wind turbines, aero engines, ships, and automobiles, and play an important supporting role in national economic development and national defense construction. Rolling bearings, as moving supporting parts, work for a long time under harsh working conditions such as high temperature, high speed, heavy load, etc., which can easily induce failures of different forms and degrees of severity, such as ball wear, metal peeling off the surface of the inner ring and outer ring raceway, retention If the frame breaks, etc., it will cause abnormal vibration of the mechanical equipment, hinder the normal operation of the mechanical equipment, reduce w...

AI Technical Summary

Problems solved by technology

At present, the methods for kernel parameter optimization mainly include grid-cross-validation method, genetic algorithm, and particle swarm optimization algorithm, but the optimization efficiency of these methods is not ideal

[0004] The Ant Lion Optimizer (ALO) is a new type of bionic intelligent algorithm proposed by the Australian scholar Seyedali Mirjalili. It has been successfully applied to the optimization of pole system structure, antenna layout optimization, site selection of distributed systems, and distribution of useless work. Electrical problems, community mining in complex networks, etc., but less research has been done in bearing fault diagnosis

In the classic

Images