A Heuristic Algorithm Based on Temporal Conflict Network for Agile Earth Observing Satellite Scheduling Problem

Agile Earth Observing Satellite (AEOS) scheduling problem consists of selecting a subset of tasks and developing observation plans for a set of agile satellite resources with the purpose of maximizing the total reward of arranged mission observations. This problem has attracted much attention in recent years since AEOS is a new generation satellite being developed all over the world. Due to its NP-hardness, heuristic methods are widely adopted when solving the AEOS scheduling problem (AEOSSP). In this paper, we propose a temporal conflict network-based heuristic algorithm (TBHA), for AEOSSP. The novelty of TBHA lies in the fact that the heuristics are extracted from a temporal conflict network, which characterizes the overlaps (conflicts) of the visible time windows of the problem. These heuristics are highly effective since they well address the time window conflicts which otherwise pose a significant challenge on the choice of the imaging start time for satellite observations. The extensive simulation experiments with the comparison to a number of heuristic algorithm variants and sophisticated meta-heuristic algorithms are conducted to show that the TBHA algorithm performs very well in terms of both solution quality and computational efficiency.