基于深度聚类网络的电子设备故障检测方法

Electronic Equipment Fault Detection Method Based on Deep Clustering Network

对基于深度聚类网络的电子设备故障检测方法进行了研究,并设计了一种基于深度聚类网络的电子设备故障检测系统。首先,对电子设备故障检测方法的总体方案进行设计;然后为了进一步提高系统对故障检测的准确率,在深度聚类网络中引入自注意力机制与残差块对其进行改进与优化;最后对优化后深度聚类网络搭建的电子设备故障检测系统进行实验测试。实验结果表明:经过引入自注意力机制和残差块后的深度聚类网络优化算法的聚类准确率达到了94.28%,比传统深度聚类网络算法的准确率提高了4.17%,CH指数与轮廓指数有明显提升,戴维森堡丁指数降低了8.06%,代表聚类效果在算法得到优化后变得更佳;采用优化后的深度聚类网络搭建的电子设备检测系统的故障诊断准确率达到了93.91%,比传统聚类算法提高了24.03%,比基于降维的传统聚类算法提高了6.26%,表明采用的优化后的深度聚类网络在电子设备故障检测方面具有明显的优势,基于深度聚类网络的电子设备故障检测系统具有可行性与有效性,且故障诊断准确率较高。

This article studies the fault detection method for electronic devices based on deep clustering networks,and designs a fault detection system for electronic devices based on deep clustering networks.Firstly,design the overall scheme of electronic equipment fault detection methods;Then,in order to further improve the accuracy of fault detection in the system,a self attention mechanism and residual blocks are introduced into the deep clustering network to improve and optimize it;Finally,experimental testing was conducted on the electronic device fault detection system constructed by the optimized deep clustering network.The experimental results show that the clustering accuracy of the deep clustering network optimization algorithm after introducing self attention mechanism and residual blocks reaches 94.28%,which is 4.17%higher than the traditional deep clustering network algorithm.The CH index and contour index are significantly improved,and the Davidson Boding index is reduced by 8.06%,indicating that the clustering effect becomes better after the algorithm is optimized;The fault diagnosis accuracy of the electronic device detection system built using the optimized deep clustering network reached 93.91%,which is 24.03%higher than traditional clustering algorithms and 6.26%higher than traditional clustering algorithms based on dimensionality reduction.This indicates that the optimized deep clustering network used in this article has obvious advantages in electronic device fault detection,and the electronic device fault detection system based on deep clustering network is feasible and effective,And the accuracy of fault diagnosis is high.