The Research About The Bearing Life Prediction And Reliability Analysis

Along with modern technology progressive and sophisticated technology developmental, the application of high quality and redundant reliable bearings day by day is also widespread, especially in the field of aerospace have been widely applied. When doing test about bearings life in the wake of it, the small sample datas and Non-failure phenomenon will occur, therefore bearings life forecast is difficult to forecast. To be able to accurately predict the life and reliability of bearings, this article used separately the Weibull distribution theory and the Fuzzy theory for studying the bearings life and reliability.This article introduces the bayesian statistical origin and development , and analyzes of test data methods of statistical inference, and the bayesian methods are discussed in the prior distribution, the likelihood function and the posterior distribution. Based on Bayesian statistical theory, building a mathematical model of the cumulative failure probability, and based on least-squares method for solving a two-parameter Weibull distribution of the two unknown parameters, getting the reliability of mathematical models of bearing. Based on the three parameters weibull distribution theory, analyzing weibull distribution three parameters of geometry and physics meaning, the paper builds weibull likelihood equations, based on Newton iterative method, and getting three parameters weibull distribution bearing reliability mathematics model.By analyzing the factors affecting bearing life, namely the establishment of the factors set, the alternative set and the weight set, using the weighted average method to calculate bearing life blurred. By analyzing we can see that in the computational process,because considering the influence life factors, and using various factors degree of membership value, therefore the computed result and the actual life are closer.Analysis and comparison two parameters of weibull distribution statistics and three parameters weibull method for the same group of bearing life through data processing, when the reliability of the bearing is in [0.9, 1], using two parameters weibull distribution and three parameters weibull distribution calculating bearing reliability has difference, but when the reliability is less 0.9, calculating results are approximate.