Improving Autonomous Underwater Vehicle Navigation: Hybrid Swarm Intelligence for Dynamic Marine Environment Path-finding

Abstract

Underwater research and monitoring operations rely significantly on Autonomous Underwater Vehicles (AUVs) for scientific investigations, resource management, and monitoring, and underwater infrastructure is provided maintenance levels amid other applications. Efficient navigation and preventative methods are only a couple of the numerous challenges that Path-Finding (PF) in rapidly changing and sophisticated Underwater Environments (UE) requires overcoming. Dynamic environments and real-time improvements are problems for traditional models. In order to provide superior solutions for navigating uncertain UE, this work suggests a hybrid optimization technique that combines Ant Colony Optimization (ACO) for local path selection with Particle Swarm Optimization (PSO) for global path scheduling. Runtime efficiency, accuracy, and distance focused on decrease are three metrics that demonstrate how the PSO-ACO hybrid method outperforms conventional algorithms, proving its significance for improving AUV navigation. The improvement of AUV functions in fields such as underwater research, along with others, is supported by the current research, which further assists with the invention of Autonomous Underwater Navigation Systems (AUNS). The PSO+ACO hybrid method is superior to the PSO, ACO, and GA algorithms in pathfinding with a 6.43-second execution time and 93.5% accuracy—the ACO model completed in 12.53 seconds, superior to the proposed system.