Monte Carlo fixed-radius floating random walk solution for potential problems

Abstract

A new Monte Carlo method is introduced. The fixed-radius floating random walk combines the strong points of the two classical Monte Carlo methods, namely the fixed random walk and the floating random walk. The method is used to solve Poisson's equation. Two typical problems in rectangular and axisymmetrical solution regions are used to illustrate the method's lower computation time and higher degree of accuracy when compared with the fixed random walk.