Testing significance of random effects for the Wiener degradation model

This paper considers the Wiener degradation model with random effects. Random-effect models take into account unit-to-unit variability of index. It is assumed that a parameter has truncated normal distribution. During research, expression for maximum likelihood estimates and reliability function been obtained. Two statistical tests have proposed to reveal existence effects in data corresponding model. The first test well-known ratio test, second one based on variance estimate parameter. These compared terms power Monte-Carlo simulation method. result research shown criterion more powerful than case considered pairs competing hypotheses. An example analysis using turbofan engine considered. set includes measurements recorded from 18 sensors 100 engines. Before constructing model, single index obtained principal component hypothesis effect insignificance rejected both tests. random-effect describes failure time distribution accurately fixed-effect