UDC 004.023

Authors

Karpov P.M.
Averkin A.N.
Federal Research Center "Informatics and Management" of the Russian Academy of Sciences; Plekhanov Russian University of Economics, Moscow, Russia

Abstract

Present work proposes novel cooling schedules for the simulated annealing algorithm. Analytic forms of new schedules were derived based on heat capacity models and the constant thermodynamic speed condition. Computational experiments performed on multiple optimization problems demonstrate that the new schedules significantly outperform the widely employed geometric cooling.

Keywords

Combinatorial optimization, Global optimization, Simulated annealing, Metaheuristic