Decision Making Support of UAV Path Planning for Efficient Sensing in Radiation Dose Mapping

摘要

This paper introduces a UAV path planning system to provide path plans for radiation dose mapping near residential area in meter-level resolution. The proposed method first identifies void areas, i.e., the GPS coordinates without sufficient measured points within a search radius, from the observation result of the previous UAV flight. The proposed method employs the flood-fill algorithm for path planning within each chunk of adjacent void areas, and the 2-opt algorithm performs path planning between the chunks of void areas. From the result obtained in our real-world case study of radiation dose mapping near an area affected by the Fukushima Daiichi Nuclear Power Plant disaster, the proposed method significantly reduces flight time compared to the results given from the algorithms of 2-opt only and flood-fill only, respectively. More importantly, one of our findings is that, even if multi-rotor UAV is used, the number of short turns during the flight is the major factor of actual flight cost.