Collaborative Search and Target Capture of AUV Formations in Obstacle Environments

When performing cooperative search operations underwater, multi-autonomous underwater vehicles formations may encounter array-type obstacles such as gullies and bumps. To safely traverse the obstacle domain, this paper balances convergence time, transformation distance sensor network power consumption, proposes a Formation Comprehensive Cost (FCC) model to achieve collision avoidance of formations. The FCC is used instead fitness function genetic algorithm solve assignment capture positions improved neural self-organizing map (INSOM) proposed efficient path-planning during process. simulation experiments in 3D space verify that scheme can improve efficiency robot deployment while ensuring safety.