Algorithm for Performance Sensitivity Estimation of Markov Discrete Event Dynamic System

摘要

An efficient simulation algorithm for performance sensitivity estimation of Markov discrete event dynamic system is designed. This algorithm can give consistent and reliable results for both steady state and transient performance sensitivity estimation. Compared with other simulation based sensitivity estimation methods, the proposed algorithm is simple in mathematical description and is easy in program realization. The capability of this algorithm is illustrated with several examples. Furthermore, our research shows that performance sensitivity of Markov discrete event dynamic system is the sum of the estimation obtained from both smoothed perturbation analysis (SPA) and likelihood ratios method (LR), singly using any one of them usually can not give consistent and reliable result.