The search for orphan radioactive source based on GPR-PRM method

Abstract

The loss of radioactive sources poses potential risks to public safety, in order to safely recover radioactive sources, a method for orphan radioactive source localization is proposed. The method is divided into two parts: constructing the radiation field by Gaussian process regression (GPR) and path planning for the mobile robot using the probabilistic roadmap (PRM) method. A mobile robot collects data from grid cells along its path, alternating between the two parts until the source localization condition is met. The path planning is determined by the PRM method combining with the improved A* algorithm, which can also help the robot jump out of local optima. The feasibility of the proposed method is verified through two simulation experiments. The first experiment comprehensively demonstrated the search process of the proposed method. In the second experiment, a comparison of different initial parameter pa was added to show that the initial search preference of the robot has an impact on the efficiency. Overall, this method enables high-precision localization of indoor radioactive source in the presence of radiation shielding at low equipment costs.