Abstract

This paper uses Max-Min Ant Colony System (MMACS) with association of narrowing of search domain and adaptive number of search nodes in each level of algorithm execution. Since these two elements have important role in algorithm`s speed and accuracy this paper discuses about these two elements and suggests making search domain narrower by special considerations in each level of algorithm execution and respectively adapted number of search nodes with this issue. It makes the algorithm faster and led's to better convergence. The algorithm has been evaluated on standard high-dimensional benchmark functions and compared with improved PSO and Genetic algorithms respectively.