Accelerated-compensated degradation modeling based on synthesized-stress enhancement test data

Abstract

At present, the reliability enhancement test (RET) data are not used adequately in the field, and are utilized to determine the limits of stress in most cases. However, the RET data which may contain abundant information of products are obtained in complex stress environment with a duration of test time. Thus, an accelerated-compensated degradation modeling method is proposed to describe the feature of RET data which are collected under temperature step stress and vibration constant stress. The model proposed is validated by using the reliability enhancement test data of a certain type of servo system. For the small sample problem, the parameter bootstrap method is used to verify the model validity. Using the enhancement test data to carry out degradation modeling, not only makes full use of data mining information, but also lays the foundation for the subsequent product reliability assessment, shortens the development time and saves the cost of degradation test.