Multi-Vehicle Exploration and Planning in Unknown Indoor Environment

Abstract

Exploration in unknown environments plays an important role in the field of mobile robots, with multi-vehicle collaboration showcasing advantages such as parallel processing, fault tolerance, flexibility, and information redundancy. This collaborative approach exhibits immense potential in applications like extraterrestrial exploration, military reconnaissance, and post-disaster search and rescue. This paper proposes a method for cooperative exploration by multiple unmanned ground vehicles (UGVs). It employs an interactive approach based on Hgrid partitioning to ensure that all UGVs simultaneously explore distinct regions. Additionally, a CVRP formulation is introduced to optimize the coverage path in unknown spaces, effectively balancing the workload distribution among individual UGVs. In the context of task allocation, each UGV employs a two-stage exploration-relocation approach to systematically cover its assigned sub-region while dynamically navigating around obstacles in real-time. Ultimately, the efficacy and feasibility of the proposed multi-vehicle cooperative exploration system are demonstrated through both simulation and real-world tests.