Intelligent task allocation in autonomous underwater vehicle swarms using ant colony optimization and contract net protocol

Coordinating many self-driving underwater vehicles in changing underwater conditions is very difficult, especially when tasks are being shared. In this study, we will introduce ACO CNP, a method that mixes Ant Colony Optimization with a better Contract Net Protocol and a new superagent system for smart, priority-based task scheduling in groups of AUVs. Unlike the standard Contract Net Protocol (CNP) methods, which have fixed assignments rules and no built-in priority queuing, or Iterated CNP (ICNP) versions that depend on heavy communications and multiple rounds of bidding to handle agent conflicts, the pro- posed ACO CNP system changes the negotiation process. It uses a flexible super-agent setup to keep central list of capabilities while carrying out assignments in a distributed way. When multiple bids come from equally qualified candidates, the ACO CNP avoids repeated exchange of messages. Instead, it uses a natureinspired Ant Colony Optimization (ACO) method to asynchronously evaluate past performance signals and current fit of resources. This combined approach avoids the large communication load common in ICNP during busy times. Tests with up to 1920 tasks running at once show that ACO CNP cuts system failure rates by 51.5 % (from 0.068 in optimized ICNP to 0.033) by naturally prioritizing important tasks and solving resource conflicts without overloading limited acoustic communication channels. This framework provides a very scalable foundation for coordinated multi-agent work in marine environments with limited bandwidth. The method works well as the number of tasks grow and can handle both priority and new tasks easily. This research uses teamwork theory to help underwater vehicles solve real problems, such as mapping the ocean floor, checking the environment, and working together on search missions.