Fault Prognosis Method Research For Key Vulnerable Components Of Electric Sliding Plug Door

As an important subsystem of industrial field and urban rail transit,the safety and reliability of electric single throw mechanical equipment(ESTME)is very important.As a typical ESTME,the electric sliding plug door are operated frequently,which make the lead screw and the limit switch become fault-prone parts.However,the maintenance department has not worked out a reasonable repair method for these failure of the vulnerable the components.Once the damage will lead the device as a whole not to run normally.Therefore,it is very important to study the fault prediction technology of the key component of electric sliding plug door and research the active fault prevention.At present,the research on the fault diagnosis technology of electric single throw mechanical equipment is a lot of research results.However,the fault prediction technology of electric single throw mechanical equipment is still in the intial stage.The main research contents of the dissertation are as follows.1)This paper introduces the structure and working principle of electric single throw mechanical equipment,and gives the two kinds of key the vulnerable parts of ESTME.At the same time,research status and shortage of fault prediction technology are given.2)The electric sliding plug door as a typical ESTME,the two kinds of progressive failure of lead screw poor lubrication and limit switch for lack of elasticity of spring leaf are used as the research object.The test platform was set up respectively,and data acquisition was designed to collect the full life degradation data as further theoretical analysis.3)Aiming at lead screw poor lubrication,the residual useful life prediction method based on Self-Organizing Feature Map(SOM),Hidden Markov Chain(HMC)and Monte Carlo(MC)simulation were proposed in the paper.By simulating structure and motor characteristics of electric sliding plug door,and the motor current signal collected was used for the multi-parameter feature extraction.The SOM algorithm was used to achieve data fusion and coding.Then,HMC model was used to train by the result.Finally,MC simulation was used to finish the residual useful life predictionof poor lubrication.The result shows that the proposed fault prediction can effectively predict residual useful life of electric sliding plug door poor lubrication.4)Aiming at the shortage of the traditional state-based prognostic method,a novel fault prediction method state-based prognostic with duration information(SBPD)was presented in the dissertation.The duration relation of each state is considered in the method.The limit switch of ESTME is considered as the research object.To solve the problem of limit switch for lack of elasticity of spring leaf,through accelerating fatigue experiment for a limit switch.Finally,SBPD model and MCMC method were used to calculate the residual useful life of limit switch,respectively.The result shows SBPD model is superior to MCMC simulation.SBPD model can be effectively used to predict the remaining useful life of the limit switch,which is the fault of lack of elasticity of spring leaf.The summary and prospect are given at the end of whole dissertation.