UAV-BASED SEARCH SYSTEM FOR "ORPHAN" SOURCES OF IONIZING RADIATION

Abstract

This article discusses a UAV-based system used for effective search and localization of radiation sources over which control has been lost. The main factors affecting the accuracy of aerial gamma-ray imaging are considered, namely: physical parameters of the system, terrain scanning technique and data processing methodology. The system structure is shown and its main technical parameters are given. The data acquisition methodology used is aimed at ensuring the completeness of measurement information, reducing uncertainties, optimizing time and resources, and increasing accuracy when searching for radioactive sources. The role of data processing is to detect radioactive sources, identify and localize them, and obtain estimates of the concentration of radioactive contamination on the earth’s surface. Correct interpretation of measurement data allows you to distinguish the source from random background fluctuations, speed up localization, and reduce the risk of errors. Particular attention is paid to data processing, which for onboard gamma spectrometry consists of the following procedures: calibration, preprocessing (data verification, spectral smoothing, correction of "live" measurement time, background correction, spectral correction, height correction, data restoration to elemental concentrations, data alignment), mapping and contouring of radioactive contamination zones, search and identification of radioactive sources, their geographical localization. The proposed stages of pre-processing provide: increasing the reliability of measurements, reducing the influence of extraneous factors, bringing data to a single scale, spectral alignment, highlighting peaks for isotope classification, determining activity, dose rate, reducing the volume of data for further processing. GIS (QGIS, ArcGIS) are used to visualize the radiation field. The Hough transform was used for the source localization process, which provides determination of the centers of each circular distribution of radioactivity. Special attention is paid to the practical conditions of using gamma spectrometry based on UAVs to search for radioactive sources. Original algorithms are used to detect and localize the source. The proposed structure of the UAV-based system, data processing algorithms allow to localize sources with radioactive radiation in quasi-real time. To verify the validity and performance of the proposed strategies, we conducted large-scale numerical experiments with different numbers of sources and their locations. The results obtained showed that the proposed approach significantly increases the accuracy, speed and efficiency of searching for radioactive sources, and also provides the ability to control, monitor and respond promptly to threats of nuclear terrorism or emergencies at nuclear power plants.