文章基本信息

标题：Self-Organizing Maps Applied to Engine Health Diagnostics
本地全文：下载
作者：Thomas Bryant ; Michael Hodges ; Mohamed Zohdy 等
期刊名称：International Journal of Computer and Information Technology
印刷版ISSN：2279-0764
出版年度：2014
卷号：3
期号：2
页码：205
出版社：International Journal of Computer and Information Technology
摘要：This research uses an algorithm implemented by the use of a self-organizing feature map (SOFM) to analyze the similarities and differences between engine sounds to indicate their health status. A neural array feature map was subjected to variation of the parameters of the self-organizing map algorithm, improving map recognition quality. In this specific unsupervised computer learning study, we examined automobile engine sounds in varying degrees of health. With the results in the paper, we have shown how self-organizing feature maps can distinguish between engines of differing ages and health. With the results that were gathered from two tests with three sets each of engine sounds across three different domains, the clusters have shown that distinct differences between the engine sets contribute to the study of sound discrimination in a substantial way.
关键词：sound; self-organizing map; clusters; features; time; ; frequency; engine health