AI Supported Coverage Path Planning Algorithms for Seafloor Mapping

Efficient coverage path planning algorithms are crucial for unmanned underwater vehicles engaged in seafloor mapping. This article proposes a new artificial intelligence based mission planning algorithms using a genetic algorithm and particle swarm optimization to optimize vehicles trajectories. The objective function is designed to maximize mission efficiency while minimizing coverage time and energy consumption. The algorithms have been tested in a simulated environment, considering varying sonar parameters and vehicle’s physical model, as well as in real conditions during field tests. In the simulated environment, the vehicle's dynamics were modeled using the nonlinear six-degrees-of-freedom model of an underwater vehicle. Compared with manual planning, implementation of genetic and particle swarm optimization algorithms could reduce mission planning time even by 95% providing robust, energy aware solution for seabed surveys, and can be embedded in existing unmanned underwater vehicles mission planning applications.