Monte Carlo Simulation

Abstract

Monte Carlo simulation consists of imitating the stochastic behavior of a physical system. Monte Carlo simulation is often used as an alternative to analytical methods. Basic concepts of Monte Carlo simulation applied to power systems are described using an example of a system with two independent components. Random sampling, or nonsequential simulation, consists of performing random sampling over the aggregate of all possible states the system can assume during the period of interest. In sequential methods, the mathematical model of the system is made to generate an artificial history over time, and appropriate statistical inferences are drawn from this history. It is crucial to sample sufficient number of states to estimate reliability indices. The chapter describes the estimation and convergence criterion of both techniques, namely: random sampling and sequential sampling. It explains variance reduction techniques, such as importance sampling, control variate sampling, antithetic variate sampling, and Latin Hypercube Sampling (LHS).