Simulated annealing with dynamic perturbations: a methodology for optimization

Abstract

The authors review the traditional simulated annealing algorithm, presenting physical analogies of the process and a formal statement of the algorithm, and propose a modification to the classical simulated annealing process through the use of dynamic perturbation functions. This modification is shown to yield a significant improvement in performance for the 50-city traveling salesman problem. The authors suggest a methodology for mapping aerospace problems into optimization problems to be solved by simulated annealing and demonstrate this methodology on three sample aerospace engineering problems: air traffic control, aircraft mission planning, and signal path switching.<>