Modelling the behavior of social insects has attracted much research recently. Although honeybees exhibit many features that encourage their use as models for intelligent behavior, up to our knowledge, no attempt for using honey-bees as a basis for optimization has been made in the literature. Some of the features that distinguish honey-bees are division of labor, communication on the individual and group level, and cooperative behavior. This paper presents a novel search algorithm inspired by the marriage process in honey-bees. The algorithm is applied to a special class of the propositional satisfiability problems (SAT) known as 3-SAT, where each clause contains exactly three literals. Two versions of the proposed algorithm, which incorporate each a well known heuristic for SAT, are developed. The two heuristics employed for each version are GSAT and random walk. The objective of this paper is to scrutinize the algorithm and compare its behavior on 3-SAT against both heuristics alone. The analysis is undertaken using two parameters gleaned from biological concepts; these are, the colony size and the amount of time devoted for brood-care.
展开▼
