In engineering applications, the assumption of constant stress is difficult to satisfy, and the degenerated data collected in the field also reflect the volatility caused by random stress. Considering the stochastic nature of storage conditions, to make results more reliable by the field test data acquisition reliability inference, this paper proposes a contact accelerated aging test method and using the acquiescence of the variability of the Taylor series, and degradation rate of Brown motion model approximation to predict field reliability is realized. Anglicizing of degradation reliability under random stress is also realized. Through simulation research, the accelerated model proposed and its acceleration effect of products under random stress are verified, and it is proved that the proposed method is an effective tool for accelerated aging test and field application.