Abstract:A method of intelligent fault detection and diagnosis based on the Support Vector Machine (SVM) is proposed. By measuring the vibration signals of the gear system at different rotating speed for different condition and faults, the testing signals were obtained. The feature signals of system were extracted and analyzed. SVM was used for the gear fault diagnosis, the classifiers of two and multi-classification were set up, and the algorithms of two and multi-classification of SVM were discussed. After analyzing, training and testing the samples of simulation data and gear vibration signal, the various damages in different running condition for gear system were detected, classified and diagnosed. Based on these, the various representative gear damage in different condition can be distinguished very well, the detection ratios is more higher as 95% in low rotating speed, and especially the identification ratios of Multi-Faults diagnosis are over 81%. As a result, Support Vector Machine in gear fault diagnosis has well diagnostic and identify abilities and development prospects, and it is an effective new method for damage detection and fault diagnosis used in engineering.