Acoustic Emission Rail Vehicle Axles Risk Assessment Based On VAR Model

This paper is a preliminary exploration of axle acoustic emission signal vector autoregression model analysis, risk assessment and management concepts and implementation process of generalization. Due to Vector autoregression model (VAR model) isn’t based on strict economic theory,VAR model is easy and there is a lot of applications, so it achieved remarkable results. The main impact of the VAR model is to calculate the relationships between variables and forecast the trend of each variable. However, due to the need for a stable data of a VAR modeling,but acoustic emission is always no-stable data.So we neeed to make the data stable.by Loeal Mean DecomPositio (LMD). LMD decomposition of the signal components of different PF, each component of PF is stationary time series data. Then build vector autoregression model by each component PF, you need to establish a set of data models with different lag order, the purpose is to compare several models and choose the most optimal model system to reflect the characteristics of the object of study and model its prediction, feature vector extraction and pulse response analysis. These data processing steps are required to using MATLAB. With econometric toolbox in MATLAB, making the whole process very easy. Risk assessment answer three questions:1) What is wrong.2) How much the probability of error.3) What are the consequences. The first problem is the identification of risk. The second problem is about the risk modeling, measurement and quantification. The third problem is the risk rating. Risk assessment can be summarized as risk analysis and risk assessment. Risk analysis and risk assessment is a necessary condition for risk management. HHM described the picture of the system being studied, so that the associated risks and highlights the risks of each subsystem to achieve the identification of risk from different angles, perspectives.