Multi-robot cooperative search for radioactive sources based on particle swarm optimization particle filter

Abstract

Effective management and monitoring of radioactive sources are crucial to ensuring nuclear safety, human health, and the ecological environment. A multi-robot collaborative radioactive source search algorithm based on particle swarm optimization particle filters is proposed. In this algorithm, each robot operates as a mobile observation platform using the latest observations to fuse into particle sampling. At the same time, the particle swarm optimization algorithm moves the particle set to a high-likelihood area to overcome particle degradation. In addition, each particle can learn from the search history of other particles to speed up the convergence of the algorithm. Lastly, the Dynamic Window Approach (DWA) for dynamic window obstacle avoidance is used to avoid obstacles in complex mountainous terrains to achieve efficient source search. Experimental results show that the search success rate of the proposed algorithm is as high as 95%, and its average search time is only 3.43 s.