Affiliation:
1. Institute of Statistics National Tsing Hua University Hsinchu Taiwan
2. Institute of Statistical Science Academia Sinica Taipei Taiwan
Abstract
AbstractDegradation models based on heterogeneous Wiener processes are commonly used to assess information on the lifetime of highly reliable products. An optimal test plan given limited resources is generally obtained using numerical methods for heterogeneous Wiener processes. However, numerical searches for optimal test plans have the disadvantage of being time‐consuming and may provide unclear explanations for the findings. To overcome these difficulties, we derive an explicit expression for decision variables (such as the termination time, number of measurements, and sample size) of ‐ and ‐optimal test plans with cost constraints. The theoretical results not only ensure that the optimal test plan is found, but also provide clear insights into the decision variables affected by model parameters and experimental costs. Some numerical examples are presented to support the efficiency and applicability of the optimal test plans.
Funder
National Science and Technology Council
Chinese Academy of Sciences
Subject
Management Science and Operations Research,Ocean Engineering,Modeling and Simulation
Reference32 articles.
1. Almeida C. P.(2011).Methods of numerical estimation for degradation models with nonnormal random effects[Unpublished master dissertation]. Department of Statistics Federal University of Minas Gerais.
2. Optimum Experimental Designs, with SAS
3. Accelerated Life Models
4. Computing and updating the first‐passage time distribution for randomly evolving degradation signals;Bian L.;IIE Transactions,2012
5. Experimental design for a class of accelerated degradation tests;Boulanger M.;Technometrics,1994