Assessment Method of Gear Wear Condition Based on Data Mining

To improve the accuracy of gear wear condition assessment, a novel gear wear condition assessment method based on data mining technology was proposed. In this method, a spur gear pair wear test was designed to extract oil parameters and vibration parameters within the whole life cycle of the gear pair first. Then, a clustering division of gear pair wear conditions was made, based on which, the association rules set between the monitoring parameters and gear pair wear conditions was obtained. Finally, a matching algorithm of association rules was developed to recognize the gear pair wear condition, and verified using the test data extracted from the wear test. The results show that the proposed gear wear assessment method based on data mining can effectively assess the wear condition of gear pairs, and the recognition rate is about 90%.