基于MEA-BP神经网络的封盒装置滑动轴承故障诊断方法

Failure Diagnosis Method for Box-Sealing Device Sliding Bearing Based on Mind Evolutionary Algorithm and Back Propagation Neural Network

针对封盒装置滑动轴承在生产过程中故障率高、可靠性低的问题,课题组提出了一种基于思维进化算法(MEA)的BP神经网络滑动轴承故障诊断方法。该方法通过多次的趋同和异化操作,不断优化BP神经网络的初始权值和阀值,建立了基于MEA-BP神经网络的滑动轴承故障诊断模型。利用样本集训练、测试和验证MEA-BP故障诊断模型,结果表明MEA-BP故障诊断法较未经优化的BP神经网络故障诊断法优势明显,能够较好地用于封盒装置滑动轴承的故障诊断,延长滑动轴承无故障使用时间。课题组的研究可提高包装企业生产效率。

Aiming at the problem of high failure rate and low reliability of sliding bearing in box sealing device in production process,a failure diagnosis method of sliding bearing based on BP neural network of thought evolutionary algorithm(MEA)was proposed.This method optimized the initial weights and thresholds of BP neural network through many convergence and divergence operations,and established a sliding bearing fault diagnosis model based on MEA-BP neural network.Using sample set training,testing and validating the MEA-BP failure diagnosis model,the results show that the MEA-BP failure diagnosis method has obvious advantages over the non-optimized BP neural network failure diagnosis method.It can be better used for the failure diagnosis of box-sealing device sliding bearing,prolong the faultfree service time of sliding bearings.This research can improve the production efficiency of packaging enterprises.