摘要 |
PROBLEM TO BE SOLVED: To detect its occurrence concerning abnormality out of assumption, by detecting kinds of abnormalities concerning the assumed abnormality in a system for detecting a truck abnormality of a railway vehicle analyzing measured sounds, by attaching microphones on the truck of the railway vehicle. SOLUTION: The system for detecting the railway vehicle truck abnormality comprises analyzing the measured sounds into a time series frequency component by frequency analysis; classifying the analyzed data into a plurality of categories, by clustering technique summarizing similar data by distance with the representative points on multi-dimensional coordinate space; storing to previously make data learn, at normal times and data at abnormal times at the representative points of each category; detecting the abnormality, by referring stored categories concerning data classified in each category; and deciding that an unexpected abnormality has ocurred, when being classified into non-stored categories. COPYRIGHT: (C)2008,JPO&INPIT
|