Citation

Gong P, Tang XB, Huang X, Wang P, Wen LS, Zhu XX, Zhou C. Appl. Radiat. Isot. 2019; 150: 1-13.

Affiliation

Jiangsu Nuclear and Radiation Safety Supervision and Management Center, Nanjing, 210019, China.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.apradiso.2019.04.037

PMID

31108333

Abstract

Locating lost radioactive sources quickly, precisely, and safely is very important in emergency responses of lost radioactive source accidents. This paper describes a source localization approach using an independently developed unmanned aerial vehicle (UAV) radiation monitoring system, which uses a specialized source localization algorithm. Once a radiation anomaly spot is found on the ground, an L×L (m²) square area around the anomaly spot defined as suspicious region is selected to perform an accurate source localization. Then, the UAV radiation monitoring instrument is dispatched to hover at some scheduled detection positions within the suspicious region for radiation measurements. After the last hover finished, the actual source position is calculated by the source localization algorithm program in real time. The source localization algorithm was developed on the basis of the inverse-square law and statistical methods. Five critical factors of the algorithm that may lead to errors in localization such as the meshing number in calculations, the size of the suspicious region, the number of the detection positions, the distribution of the detection positions, and the coverage range of the detection positions were studied by using measurement data from Monte Carlo simulations. Subsequently, the approach was experimentally verified for a 3.7 × 10⁷ Bq ¹³¹I source localization. Three experimental scenes were applied such as the source on the grass, next to a tree, and in a puddle. Different distributions of the detection positions and different numbers of the detection positions were studied. The best localization distance error was 30 cm within a 10 × 10 m² suspicious region, and the calculation time was not more than 0.1 s after a total survey flight of 5 min.

Copyright © 2019 Elsevier Ltd. All rights reserved.

Language: en

Keywords

Lost source accidents; Radiation detection; Radiation monitoring system; Source localization; UAV