Improving pelican optimization algorithm for solving integer and mixed integer optimization problems

Abstract

Mixed-integer linear programming (MILP) problems, prevalent in logistics, scheduling, and resource allocation, pose significant computational challenges due to their NP-hard nature, prompting the need for efficient metaheuristic approaches. This study proposes an enhanced pelican optimization algorithm (POA) that adapts the original population-based method inspired by pelican hunting strategies of prey approach (exploration) and surface winging (exploitation) for discrete domains via novel tent-shaped transfer functions. These functions enable seamless discretization of continuous solutions into binary and integer variables, offering computational simplicity and gradual curvature superior to traditional S- and V-shaped transfers. Tent-shaped functions outperformed others, yielding lowest standard deviation values and means closest to optima, demonstrating superior stability and precision for practical optimization.