Feature selection is one of keyissues in synthesis application of the oil monitoring technologies.According to the characteristics of oil monitoring information,K mean particle swarm clustering was used to realize samples unsupervised clustering.The defined feature contribution degree was calculated and used as the basis of feature selection,and the unsupervised feature selection based filter for the oil monitoring information was realized.The algorithm was applied in the information of atomic emission spectrum and FTIR spectrum gained from one marine diese〖JP5〗l’s〖JP〗 lubricant.The result indicates that the proposed algorithm can well realizes the feature selection of oil monitoring information,and reduce the number of the parameters,and avoid the problem that the important information is possibly pruned because of the existence of the characteristics of high relevance objective in oil monitoring information.