研究目的
To propose a general data-driven solution framework for imaging satellite mission planning in the Internet of Things, called imaging satellite mission planning framework (ISMPF), to address the lack of standardized and generalized solutions in the field.
研究成果
The ISMPF framework provides a standardized and generalized approach for imaging satellite mission planning, integrating MEC for efficient task assignment and machine learning for initial solutions. Experimental results show that appropriate algorithm selection can improve task revenue and completion rates, but further research is needed for model-algorithm combinations and software development.
研究不足
The framework's effectiveness is demonstrated through simulations and numerical examples, but real-world application constraints such as communication delays, hardware limitations, and environmental factors are not fully addressed. The choice of algorithms may need optimization for specific problem types.
1：Experimental Design and Method Selection:
The study involves designing a framework (ISMPF) with modules for task assignment using mobile edge computing (MEC), planning and scheduling with machine learning and heuristic algorithms, and task execution. Numerical examples and simulations are used for verification.
2：Sample Selection and Data Sources:
Test examples are designed for measurement and control and data downlink missions, with task sizes ranging from 10 to 300, involving multiple satellites and ground stations.
3：List of Experimental Equipment and Materials:
Not explicitly mentioned; involves hypothetical scenarios with satellites (LEO, MEO, HEO), ground stations, and MEC servers.
4：Experimental Procedures and Operational Workflow:
For task assignment, Algorithm-1 is applied to assign tasks to MEC servers or mobile terminals based on resource consumption and emergency status. For planning, BP neural networks and genetic algorithms are used in simulations to generate and evaluate planning schemes.
5：Data Analysis Methods:
Performance is evaluated based on profit and task completion rate metrics, with results averaged and compared across different heuristic algorithms.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
