Abstract:
Many underwater survey applications such as bathymetry archaeological, mine-counter-measure surveys, etc., can be efficiently performed using autonomous underwater vehicl...Show MoreMetadata
Abstract:
Many underwater survey applications such as bathymetry archaeological, mine-counter-measure surveys, etc., can be efficiently performed using autonomous underwater vehicles (AUVs). In all the mentioned survey applications, a region of interest needs to be covered using some sensors. Low cost AUVs have shown promise to perform these applications quickly and with relatively low mission cost. However, these AUVs are affected by the currents present in the region. Currents can help (resp. hinder) in propelling the AUV saving (resp. expending) time and energy and also can hinder with increased endurance cost. During such a process, the results mission time may be reduced significantly affecting the complete mission. If this issue is discovered after deployment then the cost of the mission increases. Additionally, certain areas may need a AUV with specific sensor to visit. For example, AUVs using 3D imaging sonar to map a particular area of interest (archaeology). We term such a constrain as a functional constraint. Further, the total mission time can be reduced significantly by introducing multiple AUVs. Therefore, there is a need develop mission path planners that take current information into account and the functional constraints for multi-AUV deployment. In this paper, we determine paths taking functional constraints (FC) and current information into account. We formulate the problem as a linear program, however, finding routes with the vehicle-target constraints is an NP-Hard problem. Due to the computational complexity we use insertion heuristic methods to determine a fast and simple paths. We also allow for the possibility of multi-depots (base stations for deploy-ments/retrieval) in the problem formulation. The developed insertion heuristics has two stages. The first stage involves determining a feasible solution and in the second stage we perform an iterative search for load balancing which implicitly minimizes the maximum route for the given set of vehicles. Fi...
Date of Conference: 30 September 2020 - 02 October 2020
Date Added to IEEE Xplore: 30 November 2020
ISBN Information: