Research On Fault Diagnosis Of Marine Gearbox Based On Variational Mode Decomposition

In the recent years,with the development of our country's marine strategy-to the deep blue,the importance of ship which is an important mean of transportation on the sea is increasing.So,higher requirements for the reliability of ship equipment has been put forward.Gear box as an important transmission device of marine mechanical equipment,its running state has become an important index of the whole machine.If the running state of the gear box can be grasped and the fault type of the gear box can be judged effectively,we can reduce the occurrence of major accidents and unnecessary losses.It also can provide certain theoretical guidance for the later maintenance.Therefore,the fault diagnosis of marine gearbox has become one of the most important research directions.In this paper,the gearbox is the research object.The fault diagnosis of gear is made by the method of variational mode decomposition(VMD)and deep brief network.The main contents of this paper are as follows:1.This paper introduces the common faults,causes of the gear and vibration mechanism of gear.Variational mode decomposition method is introduced in order to solve the problem of empirical mode decomposition(EMD)and ensemble empirical mode decomposition(EEMD).According to the principle and steps of VMD,the characteristics of VMD are summarized.And through the decomposition of the simulation signal,the results show that the effect of VMD is better than EMD and EEMD.2.The vibration data acquisition system based on LABVIEW is designed.The gear fault simulation experiment is carried out on the gearbox experiment platform,and large data are collected to provide data support for the following research.3.According to the analysis of the influence of the parameters of VMD(the number of modes and the two penalty factor)on the decomposition results,a more reasonable VMD parameter selection method is proposed.VMD is used to decompose the gear fault signal.We get the characteristics of the mode of the gear under different faults.Besides,the physical meaning of modal characteristics is analyzed.4.This paper also present a method for identifying the state of the gear based on the deep brief network(DBN)and the variational mode decomposition(VMD).It is proved by the research that the VMD can effectively denoise the original vibration signal,extract the frequency spectrum feature,and make the spectrum feature clear.It also can make the data features easier to learn by DBN The deep brief network can learn the spectrum characteristics,and classify the gear state data.The combination of the two methods not only can automatically identify the fault type of gear,but also improve the recognition rate of gear fault.