Reliability growth estimation with missing data. II

Abstract

In the analysis of a set of reliability growth data, it may happen that the failure information reported over some interval of test time is determined to be incorrect. This situation may take the form of too many or too few failures being reported. If these data are used in the analysis, the result may be distorted estimates of the growth rate and current system reliability. A practical reliability growth estimation procedure is discussed that assumes that the actual failure history over the problem interval is known. Maximum likelihood estimation procedures with missing data are given for the parameters of the US Army Material Systems Analysis Activity (AMSAA) model. The application of these techniques is illustrated by numerical examples.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>