Method For Determination Of Environmental Factors Based On Response Theory Model

Environment Data Synthesis method is one of effective ways to solve small-sample reliability estimation problem for high-value system, and this method's key is how to estimate the environmental factor. Without taking into account the interior relationship between environment stress and product's failure rate, traditional method used to estimate environment factor not only make use of less test data, but also have poor estimating performance and can not forecast environment factor, thus traditional methods' engineering application is limited.By introducing the Reaction Rate Model into the area of environment factor, A new method to estimate environment factor based on the Reaction Rate Model is put forward in this paper, which studies environment factor from the view of model analysis. Based on the Reaction Rate Model, the general formulas for environmental factor are deduced, and the physical meanings of the formula parameters are discussed. Then two research approaches to estimate environmental factor based on Reaction Rate Model are put forward. By integrating the product environment test data with theory model and using the regression analysis method, the point estimation and interval estimation formulas for environment factor based on the Reaction Rate Model are deduced. Based on the solving of key problems such as creating exponential censored sample, establishing point estimation performance criterion, interval estimation performance criterion and stability criterion, the factors which influence the estimate performance of environmental factor, the traditional method and model-based method's estimate performance and stability are completely compared respectively under failure-censored, time-censored, single stress and compound stress cases by means of Monte-Carlo simulation. Simulating result indicates that the estimation performance and stability of environmental factor using model-based method are better than those of traditional method. Finally, we give out a example in which the model-based method has been applied to estimate the environmental factors of a high-value system's modules in different vibration tests.The model-based method can not only improve environment factor's estimation performance and stability, but also cater to high-value systems' character that they usually have multi-resource test data. The model-based method will helpful to solve the small-sample reliability estimation problem, which is widely existed in high-value system, by the way of Environment Data Synthesis.