Abstract: With the widespread use of radioactive sources in various fields, incidents of loss of radioactive sources occur from time to time aiming at the problem of searching for a single radioactive sources, a particle filtering method based on the angle of arrival was proposed to improve the computational efficiency. The particle set was established to describe the position and intensity of the radioactive source, the angle of arrival was introduced into the particle update process, and the updated particles were selected according to the relative angle of the particles. Next, the following steps include: complete normalization, resampling and adaptive update. Establishing an observation vector set, whenever a new environmental observation data is measured, the optimal estimated position of the radioactive source reflected by all the current observation vectors can be obtained, and the search path can be guided closer to the position of the radioactive source, and when the particle filter meets the convergence conditions, the particles are weighted to obtain the final estimated position of the radioactive source. Simulation experiments show that this method can be applied to source seeking on mobile platform, its efficiency is higher than the existing particle filter method, and its detection area also significantly increases.