Fuzzy Control to Accelerate Numerical Simulation of Heat Transfer and Fluid Flow

摘要

Numerical simulation of heat transfer and fluid flow usually uses iterative methods to solve governing equations. Such methods widely use acceleration factor or relaxation factor #omega# to obtain solutions faster or to prevent divergence. IHI has constructed a fuzzy control system to optimize #omega# for better convergence. When equations of heat conduction by SOR method were solved with the system, for any starting value of acceleration factor #omega# the number of iterations to converge is as small as the number for optimum #omega#. In the next step, IHI improved the fuzzy rules for computations of fluid flow. The system was embeded in PHOENICS, and a stable convergence process was achieved.