Optimal Design of Accelerated Life Test with Multiple-Crossed Step- down Stresses Base on Monte-Carlo Simulation

In order to shorten test time and realize rapid evaluation of the reliability of products, a method of optimal design on multiple-crossed step-down stress accelerated life test (hereinafter written as MCSDS-ALT) based on Monte-Carlo simulation is proposed. The optimal plan of combined stresses in the accelerated life test is designed. The analogue simulation on the step-down accelerated life test with Monte-Carlo method is carried out, the asymptotic variance estimation of the product’s lifetime distribution with the stress under normal condition is regarded as the goal function, we take every test stress level and censored data under corresponding stress as variables of the design, and the related statistical analysis is also made by applying the theory of maximum likelihood estimation (MLE). The optimal model of MCSDS-ALT based on simulation is put forward. Through applying the method of cubic spline interpolation and fitting theory, the simulation scale is reduced and the test efficiency is also improved. Thus the paper can give the technical support on optimal design of accelerated test which can be applied in the life forecast of electronic equipment. Key-Words: Multiple-crossed step-down stress; accelerated life test; cubic spline interpolation and fitting; optimal design