Remaining Useful Life Of Planetary Gear Based On FDA And SVDD

Planetary gear is the key equipment in the transmission system,when it is degenerated or malfunctioning,it can cause the fluctuation of transmission or even failure.Active maintenance can ensure the reliability and economy of the equipment,and the prediction of residual life is the key research.Aiming at the complex and time-varying characteristics of planetary gear vibration data,the establishment of functional vibration data,classifier of fault location,degenerate trajectory,the accuracy of the prediction and the effect of the active maintenance of the equipment will be directly determined.Firstly,the method of function conversion of planetary gear vibration data is proposed.The data of the normal operation of the planetary gear can be converted by the Fourier basis function.In view of the fact that traditional basis functions can not be adapted to the complex variable and amplitude-frequency modulation characteristics of the planetary gear fault data,the transformation basis function model for different fault positions of planetary gears is established.The feasibility of the basis function is verified by the experimental data,and the normal operation of the planetary gear and the positions of different faults can be identified according to the R-square.Secondly,the multi-target recognition method of SVDD of planetary gear is proposed.In view of the shortcomings of the traditional correct rate,the evaluation function of the ROC curve principle is the optimization objective.The optimization effect and time of several intelligent optimization algorithms on SVDD model parameters are compared.The SVDD multi-target identification process for the different faults of the planetary gear is proposed to distinguish the possible effects of the different faults on the prediction of remaining useful life.Thirdly,the remaining useful life prediction method of SVDD of planetary gear is proposed.The SVDD model for all lifecycle data is established,the SVDD time-radius degradation trajectory is established based on the data of different running time.In view of the eccentricity of the SVDD model,the time-radius degradation trajectory of the multi SVDD model is established.The experimental sample can show that the prediction accuracy can be improved with the increase of the number of SVDD models.Finally,the planetary gear vibration test stand is set up to validate this paper method.And the programs of simulation failure and the full lifecycle test of the planetary gear are designed.Functional transformation of the collected planetary gear vibration data is completed,the function coefficients and error parameters are distinguished by optimized SVDD,remaining useful life is predicted by the SVDD time-radius degradation trajectory of full lifecycle data.The test samples have proved that this method has high prediction accuracy,strong applicability and easy to popularize and apply.