A method of distributed multi-satellite mission scheduling based on improved contract net protocol

Abstract

With the increasing number of earth observing satellites and the growing demand for remote sensing data, satellite mission scheduling is undergoing a change from the traditional off-line mode to on-line mode. To overcome difficulties in handling large quantities of satellites and large-scale scenes by the current on-line multi-satellite mission scheduling method, we establish the mathematical model, design the computation framework based on multi-agent system contract net protocol.Then we put forward a load reduction method based on mean shift clustering for satellite scheduling agent and a bid evaluation method based on genetic algorithm for central cooperating agent.Finally, Experimental results are used to demonstrate the feasibility and effectiveness of the algorithms.

Task Planning for Multiple-Satellite Space-Situational-Awareness Systems

“…Traditionally, the CNP requires heavy communication among agents to achieve dynamic task planning. In recent years, researchers developed a variety of remedies and enhancements to reduce the tender space in order to reduce communication and computation burden [16,[22][23][24]. For example, a clustering method was proposed to reduce the number of bids, thereby reducing the computational burden [23].…”

“…In recent years, researchers developed a variety of remedies and enhancements to reduce the tender space in order to reduce communication and computation burden [16,[22][23][24]. For example, a clustering method was proposed to reduce the number of bids, thereby reducing the computational burden [23]. In [24], machine-learning methods were employed to quickly initialize a feasible solution of a task-assignment problem of image-sensing satellites.…”