| Being a type of the most familiar low-voltage apparatus, electromagnetic relay is used in national defence, military affairs, automatization and other domains widely. The reliability of electromagnetic relay works on the reliability level of the whole system directly. The product studied in the article is space closed electromagnetic relay, its period of storage is by far longer than usage time. All the more, we care the reliability in its storage moment, and usually make use of accelerated storage life testing to research the reliability.When the accelerated storage life testing of electromagnetic relay is finished, the testing data which we have gotten often can not reflect the storage life of the product directly. Thus, we must make use of appropriate data processing method to analysis the testing data. Then, we may calculate actual figures of storage life for the product.This article shows five data analysis methods, which are simple linear regression, exponential regression, weighted exponential regression, polynomial regression (method of least squares) and grey theory. We make use of the five methods to fit and forecast five test parameters separately.1. At first, the article illustrates the theory of each regression analysis method. Then, the model parameters are determined.2. According to the calculating processes for the model parameters, the program of data analysis is introduced, which can accomplish data processing.3. Then, the fitting and forecasting results of the five methods are showed. At the same time, the accuracy analysis for the results is made. In the end, the article shows the comparison for the forecasting effectiveness of the five methods and storage expiration date of the product.The article makes use of VC++6.0 to program, and origin 6.0 to draw system of coordinates.This project has relation with the accelerated storage life testing of Electrical Institute in Hebei University of Technology.
|
PDF Full Text Request