Research On Prediction Of Mechanical Equipment Remaining Useful Life Based On Stochastic Process Modeling

Mechanical equipment(such as wind turbines,high-speed trains,etc.)suffers from frequent failures due to harsh working conditions and complicated working conditions,resulting in huge financial and material losses.It is of great significance to reduce maintenance costs,ensure safe,stable and efficient operation of the equipment,and avoid major losses caused by failures by timely assessment of the current health status of the equipment and predict its future working status,formulate relevant maintenance strategies based on the operational status of the equipment.In this paper,a method based on stochastic process modeling technique is used to predict the remaining useful life(RUL)of equipment.Chapter 1 introduces the development status of RUL prediction technology and the application of stochastic process in RUL prediction method.Chapter 2 constructs the health indicator(HI)of equipment when there is only one HI.The characteristics of the HI in the degradation process of equipment are analyzed.And then degradation model based on Wiener or Gamma process was built to predict the RUL.The performance of the proposed approach is vilified by using simulation and experimental data.Chapter 3 describes the construction of two-dimension degradation model based on stochastic process when there are two HIs.The appropriate Copula function,which is selected out by using the AIC rule,is used to analyze the correlation characteristics between the two HIs.The model parameters are updated by using online data and maximum likelihood estimation method,and then the RUL of equipment is predicted.Chapter 4 discusses the development of high-dimension(greater than or equal to 3)degradation model.Vine Copula function is introduced to describe the correlation characteristics between various HIs,and the RUL of the equipment based on the multivariate stochastic process degradation model is predicted.The simulation and experimental data are used to verify the above methods.Results show that the proposed method achieves better performance in predicting the RUL of the equipment.In this paper,RUL prediction approaches based on unary,binary and multivariate stochastic process modeling are discussed respectively.The Copula function and model parameter estimation algorithm are used to evaluate the current health status of the equipment and predict its RUL.The effectiveness of the proposed method is verified by simulation and experimental data.