Performance analysis and feasibility study of ant colony optimization, particle swarm optimization and cuckoo search algorithms for inverse heat transfer problems

Abstract

In the present work, three recently developed metaheuristic algorithms (ant colony optimization, cuckoo search and particle swarm optimization) are examined for a class of heat transfer problems. Unknown boundary heat fluxes are estimated for conduction, convection and coupled conduction–radiation problems. Direct problems are solved to determine temperature distribution assuming known boundary heat flux. Inverse method is then used to estimate boundary heat flux with the help of the temperature previously determined from the direct problem. To replicate experimental error, effect of noise on temperature data is introduced to examine the robustness of all the algorithms. Effect of time step size and regularization are studied. It is found that all the algorithms are promising and can be used for this class of inverse heat transfer problems. Performance of all the algorithms is comparable. Efficiency of the three algorithms is compared in terms of CPU time. Ant colony optimization algorithm is found to be most efficient followed by particle swarm optimization and cuckoo search algorithms for all the considered heat transfer problems. All the algorithms are also applied to estimate diffusion coefficient of a food material (mushroom) using experimental data.