Research On Fault Identification And Diagnosis Of Low-speed Worm Gear Reducer

Gear reducer is widely used in power transmission mechanical structure,having seriously affected the personal safety and economic benefits in industrial production.This paper used RV series low speed worm gear reducer to carry out a reducer fault identification and intelligent diagnosis research.The main contents are:1)The main fault types of low speed worm gear reducer are analyzed.A set of gearbox fault diagnosis program is designed and a speed reducer experiment platform is completed.2)The experimental data of reducer are analyzed by different methods,including analysis in time domain,frequency domain and time-frequency domain,which show that empirical mode decomposition performs better in picking up fault features.3)The SVM’s classification accuracy with different kernel function and parameters is compared,then particle swarm optimization is proposed to be used to optimize the model parameters of the classifier.The experiment result showed that the optimized classifier has higher classification accuracy then the classifier without parameter optimization.4)To counter the problem that the measurement data of a single sensor is not enough to reflect the complicated fault information of the reducer,this paper introduces the multi-sensor data fusion technology.And proposed a fusion diagnosis system based on SVM and DS evidence theory,using the output of SVM after Platt scaling as the BPA output of DS evidence theory.Experimental analysis shows that the classification accuracy of the diagnostic system after fusion has been improved.