A Parallel Intelligent Search and Rescue System for Swarm Robots Based on Digital Twin

In recent years, urban search and rescue (USAR) tasks in unknown environments have arisen widespread attention in the domain of Smart City. However, the complexity and unpredictability of unknown environments pose great difficulties for decision-making in search and rescue phases, ultimately resulting in serious consequences. To address the decision-making problems, this paper proposes a hierarchical swarm robots digital twin framework, which includes a physical layer, virtual layer, twin data layer, parallel decision layer, and service layer, to provide efficient decision support for USAR tasks in large static unknown indoor environments. Besides, considering the requirements of reality, we specifically design two modules: (1) real-time twin model modeling module, solving the issue of poor real-time performance in the offline construction of twin models. (2) parallel decision path planning module based on deep reinforcement learning, solving the difficulty users face in obtaining direct decision support. In our experiment, we implement the proposed framework using Unity3D and the Robot Operating System (ROS). The results show that the framework is capable of improving the decision efficiency of USAR tasks. Additionally, the update frequency of the proposed framework is kept in less than 1 s while providing optimal path planning.