Optimizing the Modified Lam Annealing Schedule

Abstract

Simulated annealing is a metaheuristic commonly used for combinatorial optimization in many industrial applications. Its runtime behavior is controlled by an algorithmic component known as the annealing schedule. The classic annealing schedules have control parameters that must be set or tuned ahead of time. Adaptive annealing schedules, such as the Modified Lam, are parameter-free and self-adapt during runtime. However, they are also more complex than the classic alternatives, leading to more time per iteration. In this paper, we present an optimized variant of Modified Lam annealing, and experimentally demonstrate the potential significant impact on runtime performance of carefully optimizing the annealing schedule. Received on 07 October 2020; accepted on 03 December 2020; published on 16 December 2020