Global optimization of grillages using Simulated Annealing and high performance computing

Abstract

The aim is to investigate ways of increasing the efficiency of grillage optimization. Following this general aim, two well-known optimization methods, namely the Genetic Algorithm (GA) and Simulated Annealing (SA), were compared using some standard medium size (10 and 15 piles) examples. The objective function was the maximal vertical reactive force at a support. Coordinates of piles were optimization variables. SA wins and was applied to real-life problem (55 piles) by parallel computations performed using a powerful cluster. New element is comparison of SA with GA and application of SA to a practical problem of grillage optimization.