Year: 2026
Type: preprint
Abstract: Path planning for multi-robot systems under complex task-logical specifications and strict temporal constraints poses significant combinatorial challenges. Existing approaches often handle logical and temporal requirements separately, limiting their scalability and practical applicability. To overcome this limitation, we propose a unified planning framework that seamlessly integrates Boolean specifications with time-window constraints. The problem is first formalized as a Mixed-Integer Linear Programming model that captures logical dependencies and timing requirements, including a collision avoidance module. To improve scalability for large-scale instances, we develop a Beam Search-based heuristic algorithm tailored for handling time windows. A similarity-based filtering mechanism is further incorporated to promote solution diversity and mitigate premature convergence. Experimental results demonstrate that the proposed approach achieves a desirable trade-off between computational efficiency and solution quality, offering a robust and scalable solution for multi-robot path planning under complex temporal-logical specifications.
Source: Research Square
Authors: Siyu LIU, Syrine Bouazza, Zhou He, Dimitri Lefebvre
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7233841788
