Gear Fault Diagnosis Method based on Multi-domain Feature and Improved D-S Evidence Theory

In order to fully and accurately identify the fault category of gear，a feature space model based on multi-domain characteristic parameters such as time domain，frequency domain and energy is established. On this basis, an intelligent fault diagnosis method based on multi-domain feature and improved D-S evidence theory is proposed. Relevant feature parameters are extracted from the measured data as the diagnostic samples，and multiple evidences are constructed with the preliminary diagnosis results of particle swarm optimization support vector machine（PSO-SVM）. The experimental results verify the effectiveness of the final diagnosis results obtained by the improved D-S evidence theory in this work.