Research On Electrical Products Storage Life Analytical Method

No matter in preservation or working condition, any kind of products, includingelectrical products, has a certain lifetime. As for disposable products, preservation peiodmay be much longer than the working time. Therefore, research on the storage life is moremeaningful than the working life. Storage life is a significant factor In the study ofelectrical products’ reliability. Predicting relays’ storage time precisely can not only developthe reliability of the power system but also reduce the storage cost, Therefore, research onstorage life has the very vital significance.For the prediction and analysis of electrical products’ storage life, the main researchwork is as follows:(1) the factors causing relays failure in the condition of storage wasstudied;(2) the failure mechanisms for the relay was determined and study shows that themost important factor which caused the failure of the relay was temperature (3)theacceleration models were studied under different stresses, on this basis,the accelerate lifetest program was confirmed and the original value of the storage life was got through theaccelerated life tests;(4) the classical grey model GM(1.1) and the grey Verhulst modelwere improved in the following three aspects: the background values, the initial values, andthe consecutive neighbors. Then, a composite prediction model was established and appliedto the storage life prediction for the relay.In order to test the effectiveness of the prediction models, the error analyses andcomparison were carried out between the10kinds of models, the results indicated:(1)theprediction accuracy of the improved models based on GM(1.1) were higher thanGM(1,1);(2) the grey Verhulst model as well as the improved models based on it had a verygood fitting accuracy for the historical data,but were no match for GM(1,1) in theprediction accuracy.Finally, the comparisons were taken between improved GM(1.1) model and two kindsof regression models. The results indicated that the prediction accuracy of former modelwas better than the later. In conclusion, all of the improved models based on GM(1.1) areapplicable to the storage life prediction of relay. And the precision of prediction model issuperior.