TY - JOUR
T1 - A hybrid estimation of distribution algorithm for agile earth observing satellite task scheduling problem
AU - Ma, Chunchun
AU - Huang, Panxing
AU - Liu, Xiaoze
AU - Wu, Chu ge
AU - Xu, Rui
N1 - Publisher Copyright:
© 2025 Elsevier B.V.
PY - 2025/7
Y1 - 2025/7
N2 - Agile Earth Observing Satellites (AEOSs) represent a new generation of Earth observation satellites, widely used for various observation tasks. To efficiently utilize the visible and observing durations of the orbiting AEOS, the AEOS scheduling problem (AEOSSP) is formulated to maximize the overall observation profit while satisfying the complex operational constraints. In this paper, a hybrid Estimation of Distribution Algorithm (EDA) that incorporates three knowledge-oriented local search operators is proposed to efficiently solve the AEOSSP. The multiple multidimensional knapsack problem with conflicts (MMdKPC) is first modeled and used to formulate AEOSSP. An EDA probability model as well as its updating and sampling mechanisms, is designed to generate solutions to explore the solution space and generate potential solutions. In addition, based on the characteristics of MMdKPC, three knowledge-oriented local search operators are developed to improve the solution. Based on the benchmark instances and simulation data provided sampled from Satellite Tool Kit, the comparison simulation experiments are carried out. The results validate the effectiveness of three knowledge-oriented local search operators, respectively. Additionally, the proposed hybrid EDA performs better compared to the existing state-of-the-art algorithms in terms of overall observation profit.
AB - Agile Earth Observing Satellites (AEOSs) represent a new generation of Earth observation satellites, widely used for various observation tasks. To efficiently utilize the visible and observing durations of the orbiting AEOS, the AEOS scheduling problem (AEOSSP) is formulated to maximize the overall observation profit while satisfying the complex operational constraints. In this paper, a hybrid Estimation of Distribution Algorithm (EDA) that incorporates three knowledge-oriented local search operators is proposed to efficiently solve the AEOSSP. The multiple multidimensional knapsack problem with conflicts (MMdKPC) is first modeled and used to formulate AEOSSP. An EDA probability model as well as its updating and sampling mechanisms, is designed to generate solutions to explore the solution space and generate potential solutions. In addition, based on the characteristics of MMdKPC, three knowledge-oriented local search operators are developed to improve the solution. Based on the benchmark instances and simulation data provided sampled from Satellite Tool Kit, the comparison simulation experiments are carried out. The results validate the effectiveness of three knowledge-oriented local search operators, respectively. Additionally, the proposed hybrid EDA performs better compared to the existing state-of-the-art algorithms in terms of overall observation profit.
KW - Agile earth observing satellite task scheduling
KW - Estimation of distribution algorithm
KW - Intelligent optimization algorithm
KW - Knowledge-oriented scheduling
KW - Multiple multidimensional knapsack problem with conflicts
UR - https://www.scopus.com/pages/publications/105005067083
U2 - 10.1016/j.swevo.2025.101971
DO - 10.1016/j.swevo.2025.101971
M3 - Article
AN - SCOPUS:105005067083
SN - 2210-6502
VL - 96
JO - Swarm and Evolutionary Computation
JF - Swarm and Evolutionary Computation
M1 - 101971
ER -