Research on reliability growth models for multi-systems development testing

Reliability growth models for multi-systems are studied, the different points of the models are analyzed. For the disagreement of multi-systems reliability growth models, the converse thought way is presented for model validation. Effects of improvement strategies on failure point process are considered. On the assumption that failure point process follows the Nonhomogeous Poisson process (NHPP) during reliability growth testing for single system/product. Random point process for failure stream of multi-systems can be obtained by using the theory of computer simulation. Through the analysis on the failure stream for multi-systems, the results show that the improvement strategy has an effect on the model, and current reliability growth model of multi-systems is improper for estimating reliability growth of multi-systems. Mean time between failures (MTBF) of product will be overestimated by applying the AMSAA-BISE (army material systems analysis activity-Beijing Institute of Strength and Environment) model and be underestimated by using approximated estimation model of MEI WenHua. Furthermore, the errors of the reliability estimation by using current reliability growth model are in direct proportion to the number of testing systems. However, based on the asynchronized improvement, the estimation of the model which is presented by WANG YuYing, is closer the accurate results.