Axle Acoustic Emission Signals VAR And ARMA Prediction And With Axle Analysis Hierarchy Process

The safety of axle is an important condition for locomotive running,axle is an important component of the locomotive.Able to accurately predict the emission signal is equal to predict next time the state of the axle.Everything has a risk,in the running of locomotive if axle failure will bring immeasurable loss,so the risk assessment of the axle is very necessary.In order to complete the task of the research group,the first part of this article is the use of econometric forecasting model applied to the axle acoustic emission signals.Using VAR model and ARMA model to forecast the axle acoustic emission signal and get simulation curves.And general economic data is different, the amount of data of acoustic emission signals is bigger,when selecting the best model we faced with more choices. In order to test the prediction effect of the two models,we comparing the forecast results with actual results,the results indicate that using a single model to predict the signal is not stable.According to prediction accuracy gave two models weights,by using the weighted models to predict the acoustic emission signal contrast the forecast results with single model’s.The empirical results show that the weighted prediction precision is high and more stable.The second part of this article is using analytic hierarchy process analyzed the main faults of the axle,the forecast is applied to the fuzzy comprehensive evaluation.Drawing the axle failures hierarchy analytical diagram,on the basis of that,figure out the various parts single level sequencing.Combing two kinds of risk assessment standard and combined with the actual situation,build risk rating matrix,using ARMA model to the data modeling, forecasting and get their risk level and risk membership matrix,Finally,using the fuzzy comprehensive evaluation method, evaluate fuzzy comprehensive risk of this axle.The different with the method of fuzzy comprehensive evaluation in general is that not using the expert scoring method to get the risk membership matrix,but using the latest forecast results to get the matrix.Changing qualitative expert evaluation method into quantitative method.Make the results more convincing.In this paper, on the final risk assessment of axle is level three, it’s risk is acceptable.