An automated solver to determine ML estimates for the AMSAA discrete reliability growth model

This paper describes the use of an optimization technique known as evolutionary programming (EP) to determine the estimates of the parameters required for reliability growth determination. The EP optimization algorithm was embedded into an automated spreadsheet solver and used to solve the example problem presented by Crow (1994). The example problem produced a nonlinear response surface that is difficult to solve for hill-climbing optimizers such as the Excel Solver. The automated EP solver developed by the authors presents a practical and effective solution tool to determine the parameters in the AMSAA discrete reliability growth model.