Short-term life prediction based on field reliability data and combined life stress model

Based on the field failure mechanisms in smart electrical energy meters, field operating abnormal data and maintenance data were analyzed. Discrimination analysis for field data was screened for predicting residual life. Through thoroughgoing analysis, field failure data, failure modes, failure mechanisms and external inducing stresses as well as their features with time were analyzed. Then the degradation characteristics relating to failure or product residual life with sensitive stress were defined. Classical reliability physical stress life models were applied to establish a combined stress life model for the smart electrical energy meters with different proportions by different stresses. Using the combination life model, residual life for the product batch can be predicted. This could be used for exposing latent batch's failures, replacing problematic batches in advance. Besides, it could afford theoretical and data support for fast troubleshooting and fault priority handling.