Planning accelerated life tests under scheduled inspections for log-location-scale distributions

Abstract

This paper proposes an efficient approach to planning an ALT under scheduled inspections. We aim to simultaneously optimize stress levels, sample allocation, and inspection times for lifetimes that follow log-location-scale life distributions, including Weibull and Lognormal distributions. Such a high-dimension optimization problem is solved by a computationally efficient approach leveraging on the asymptotic equivalence between the selection of sample quantiles for parameter estimation of a location-scale distribution and the selection of the optimal inspection times during an ALT for the same purpose. A numerical example is presented to illustrate the application of the proposed approach, and a sensitivity analysis is performed to investigate the robustness of the optimal ALT plans against misspecification of planning inputs. A computer program coded in the MATLAB Graphical User Interface Design Environment is provided to make our method readily applicable in practice. © 2012 IEEE.