Convergence of the simulated annealing algorithm for continuous global optimization (Q1573995)

In this paper, the convergence behavior of a class of simulated annealing (SA) algorithms for continuous global optimization is analyzed. The SA algorithms under consideration uses a temperature-dependent probability density function to generate the random trial points from the current iteration point and an annealing time-dependent temperature updating function to reduce the value of the temperature after each random trial point is generated. The author provides a convergence analysis and promises further research including numerical experiments.