Ant colony pattern search algorithms for unconstrained and bound constrained optimization

Ant colony optimization is a class of metaheuristics which succeed in NP-hard combinational optimization problems rather than continuous optimization problems. We present and analyze a class of ant colony algorithms for unconstrained and bound constrained optimization on Rn: Ant Colony Pattern Search Algorithms (APSAs). APSAs use the ant colony framework guided by objective function heuristic pheromone to perform local searches, whereas global search is handled by pattern search algorithms. The analysis results of APSAs prove that they have a probabilistic, weak stationary point convergence theory. APSAs present interesting emergent properties as it was shown through some analytical test functions.