Interactive search algorithm: A new hybrid metaheuristic optimization algorithm

摘要

In this paper, a new hybrid optimization algorithm, called “Interactive Search Algorithm (ISA)” is proposed for the solution of the optimization problems. This algorithm modifies and combines affirmative features of two developed metaheuristic methods called Integrated Particle Swarm Optimization (iPSO) and Teaching and Learning Based Optimization (TLBO). ISA consists of two separate paradigms: (i)Trackingand (ii)Interacting. Tracking paradigm utilizes the information stored in the current agent’s memory and two other important agents, the weighted and best agents, to guide the colony. On the other hand, interacting paradigm provides a pairwise interaction between agents to share their knowledge with each other. Each agent based on its tendency factor employs one of these two paradigms in each cycle of ISA to explore the search space. Additionally, rather than conventional penalty approach, ISA utilizes the improved fly-back approach to handle problem constraints. The search capability of the proposed method is tested on the number benchmark mathematical functions and constrained mechanical design problems as the real-world examples. Consequently, the achieved numerical results demonstrate that the proposed method is competitive with other well-established metaheuristic methods.