Comparison of Optimal Designs between Constant-Stress and Step-Stress Accelerated Life Tests for Competing Risks Models under Exponential Lifetime Distributions

Abstract

Accelerated life testing (ALT) is a process of testing products by subjecting it to strict conditions in order to observe more failure data in a short time period. In this work, we compare the two-level constant-stress ALT and simple step-stress ALT based on competing risks of two failure modes when the failure modes are of independent exponential lifetime distributions. Optimal sample sizes allocation on CSALT and optimal changing time of stress levels on SSALT are considered based on V-optimality and D-optimality, respectively. Under Type-I censoring, we show that the optimal SSALT is better than the optimal CSALT based on the above optimal criteria under specific conditions. Numerical results also verify that the optimal SSALT outperforms the CSALT in general. A real example is analysed to demonstrate the performance of the optimal plans for both ALTs.