Large-scale hybrid mission scheduling for LuTan-1 satellites using sparse evolutionary algorithm

Abstract

LuTan-1 is China’s first L-band Synthetic Aperture Radar (SAR) satellite system designed for high-precision global land deformation monitoring. To address the challenges of SAR satellite observation missions characterized by extensive spatial coverage and intensive temporal conflicts, this paper presents a hybrid large-scale mission scheduling optimization method based on a revised sparse evolutionary algorithm (R-SparseEA). The method first generates meta-tasks through irregular region decomposition to minimize global overlap while comprehensively considering satellite visibility, onboard resources, and regional priorities. Subsequently, a multi-objective hybrid mission scheduling model with multiple revisit periods is established, providing an efficient and scalable framework for describing the large-scale long-term decision-making problem of SAR satellite systems. The proposed R-SparseEA algorithm incorporates innovative evolutionary techniques, including novel population initialization, masked genetic operations, and sparse population revision strategies to effectively solve this model. These techniques ensure both the feasibility and sparsity of the solution set throughout the evolutionary process. Comparative experiments demonstrate that R-SparseEA outperforms three state-of-the-art sparse evolutionary algorithms in Pareto solution set distribution, convergence performance, and computational efficiency. Simulation results indicate that complete coverage of China can be achieved within 33 days, while global land observation can be accomplished in 79 days through the collaboration of LuTan-1’s dual satellites.