Fast marching firework method for multi-goal mobile robot path planning in complex obstacle maps

Tran Thi Cam Giang, Huynh Thi Thanh Binh, Ho Viet Duc Luong, Nguyen Huy Hoang, Do Quoc Huy

Abstract: Multi-goal mobile robot path planning is a challenging optimization problem in robotics, frequently required for tasks such as surveillance missions, data collection, and active perception. The objective is to find an efficient path that avoids collisions while visiting all predefined desired goals in an environment with obstacles. The existing approaches, such as the Space-Filling Forest and the Rapidly-Exploring Random Tree, address this problem in two phases: they create a collision-free path between each pair of goals and then solve the Traveling Salesman Problem using the pre-calculated goal-to-goal distances. These methods offer the advantage of creating a distance matrix; however, they have limitations: tree expansion stops when two trees collide, potentially missing certain pathways blocked by other trees, and computation time is considerable. To address these disadvantages, this paper proposes a novel path planner called the Fast Marching Firework (FMF) algorithm. In FMF, each goal acts as a source of fireworks, expanding simultaneously from the goals and continuing to intersect when they encounter each other, creating new paths between goals through more contact points. Numerical results manifest the performance and the superiority of the proposed method on various types of maps. The results demonstrate that FMF reduces travel costs by 9.07% compared to existing algorithms and requires less computation time to cover all areas of the environment.

External IDs:dblp:journals/isrob/GiangBLHH25