Dynamic Mission Planning Performances for Agile Earth Observing Satellites with Adaptive Multi-Agent System
Year: 2025
Type: conference-paper
Abstract: Planning an Agile Earth Observation satellite constellation is complex due to the need to manage overlapping objectives, limited resources (e.g., time availability, platform agility, onboard energy and data storage), satellite heterogeneity and dynamic environmental constraints like weather or user need emergencies. The on-ground mission planning problem for agile Earth observation satellite systems is traditionally addressed using greedy algorithms (often hierarchical), approaches based on linear or constraint programming, problem-specific heuristics, and metaheuristic algorithms such as genetic algorithms or ant colony optimization. These traditional methods are well suited for static scenarios but struggle to adapt to dynamic and reactive environments with evolving constraints and real-time updates. This paper evaluates the performance of a self-Adaptive Multi-Agent System (AMAS) approach in addressing this challenge. By modeling satellites and tasks as autonomous agents, the developed system enables decentralized decision-making, improving adaptability, task allocation, and resource management. Experimentation results show that the self-Adaptive Multi-Agent System outperforms the hierarchical greedy approach, achieving higher responsiveness, better task completion rates, and reduced computational overhead, demonstrating its suitability for dynamic on-ground mission planning in real-world satellite operations.
Source: HAL
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Funder: European Space Agency
OpenAlex ID: W7153683724
