Swarm intelligence algorithms: a survey of modifications and applications

Abstract

Swarm Intelligence (SI) is a dynamic subfield of artificial intelligence that draws inspiration from the collective behaviors of natural systems ‎such as ant colonies, bird flocks, and fish schools. This paper provides a comprehensive review of SI algorithms, examining their foundational ‎principles, recent modifications, and applications across diverse domains. Prominent algorithms such as Particle Swarm Optimization (PSO), ‎Ant Colony Optimization (ACO), Artificial Bee Colony (ABC), and Bat Algorithm (BA) are analyzed alongside emerging approaches like Grey ‎Wolf Optimizer (GWO), Zebra Optimization Algorithm (ZOA), and hybrid frameworks. A key focus is placed on algorithmic advancements, in-‎cluding adaptive inertia weights in PSO, pheromone update mechanisms in ACO, and hybridization techniques such as GWO-PSO and WOA-BA, ‎addressing challenges related to convergence speed, scalability, and robustness against local optima.‎ This review explores the practical applications of SI algorithms in engineering design, healthcare, robotics, logistics, education, and social ‎media. Detailed performance comparisons reveal the strengths and limitations of each algorithm, supported by empirical results from ‎benchmark problems such as the Traveling Salesman Problem (TSP), pressure vessel design optimization, and radiotherapy planning. Addi-‎tionally, the study highlights novel algorithms developed between 2020 and 2023, shedding light on their contributions to the field. The ‎paper concludes by identifying current challenges, such as computational overhead and parameter sensitivity, and suggests future directions, ‎including the integration of machine learning, lightweight adaptations for resource-constrained environments, and bio-inspired enhance-‎ments‎.