Abstract
The rapid search for missing radiation source can reduce the harm to the human body and the generation of panic, which is of great significance to the people's life, work and social stability. This paper is based on solving the problem of searching for radiation source in a complex environment closer to reality, construct obstacle scenarios, and consider the situation of limited communication. Use the consensus-based distributed source parameter estimation algorithm to process the detection values and the posterior information, which is obtained by itself and its neighbors, and the location and activity of the missing radiation source are gradually estimated. Then the variable step length distributed free energy strategy is used to guide the robot to the next position to continue detection. The two processes run iteratively until the missing radiation source search task is completed. The sharing of neighbor information within the communication range will increase the particle diversity of the local particle filter algorithm, improve the estimation accuracy, and also achieve the consistency of source parameter estimation and source search result. Experimental results show that the algorithm in this paper has a higher search success rate and shorter search time in complex environment.
