Evaluation Method for Accelerated Degradation Testing Considering Uncertainty of Failure Threshold

摘要

The traditional accelerated degradation testing (ADT) technology usually evaluates reliability and life of products through analyzing degradation mechanism and processing degradation data with probabilistic method. However, different failure thresholds will result in different evaluation results. In traditional evaluation method of ADT, the failure threshold of product is influenced by the environmental factors and the cognitive level of experimenter, which makes the failure threshold ambiguous and affects the accuracy of evaluation. Ignoring the uncertainty of failure threshold may lead to the fact that results of ADT statistical analysis can be hardly used as the basis for life or reliability evaluation of product. Thus, it is very necessary to analyze the uncertainty of failure threshold during ADT evaluation, which can help deciders make a more credible evaluation of products. This article introduces the traditional evaluation method of ADT and analyzes the importance of the failure threshold to the evaluation result. Then, fuzzy model and membership function of the failure threshold are modeled and analyzed. After that, the reliability function considering the uncertainty of failure threshold is obtained. At last, based on evaluation method considering uncertainty of failure threshold, a case study of lithium-ion battery (Li-ion) is presented and its effectiveness is verified.