Research On Reliability Evaluation Method Of Mechanical Products Based On Test Degradation Data

With the improvement of product reliability,a large number of mechanical products with high reliability and long life are widely used.For such mechanical products,life data is difficult to obtain,so it is difficult to conduct traditional reliability assessment based on life data.The failure of mechanical products is a process of gradual degradation.It has become a research hotspot to fully explore the effective information of the degradation stage before the failure of products and to carry out reliability analysis and evaluation on the products with high reliability and long life by using the degradation model.The main research work of this thesis is as follows:First,in view of the mechanical product is not easy to get life data,degradation data exist volatility,the difference between the sample problem,this thesis puts forward a kind of considering individual differences Wiener process degraded reliability assessment model,namely in the construction of the model,not only considering the volatility,considering the amount of degradation of initial value at the same time,the drift coefficient and diffusion coefficient of randomness.According to the size of their coefficient of variation,they are set as a constant value or normal distribution.In this way,the influence of individual differences between products is fully considered to make the evaluation results more accurate.In addition,from the perspective of time series,a reliability evaluation method based on ARIMA model is proposed.ARIMA model has a good evaluation effect on the degradation data with dynamic and nonlinear characteristics,and only considers a single variable.Instead of taking the complex failure mechanism of mechanical products as the basis,it evaluates the product life with the degradation law of a series of degradation sequences,which is simple and clear.Then,starting from the reliability evaluation of degradation data of mechanical products,the life test was designed and degradation data were collected by taking high-pressure valves as an example to illustrate that the two methods could combine the basis and ideas of comparison to establish an evaluation model,and the unknown parameters of the two models were estimated by the maximum likelihood method,and the evaluation results were obtained.Finally,the results of reliability evaluation and life prediction of high-pressure valves based on Wiener process degradation model and ARIMA model are compared and compared with the real values,indicating that the results of reliability evaluation of Wiener process degradation modeling considering individual differences are more reasonable and effective.