A Real-Time Effectiveness Evaluation Method for Remote Sensing Satellite Clusters on Moving Targets.

Area of Science:

• Earth observation
• Remote sensing technology
• Satellite systems engineering

Background:

• Remote sensing satellites are crucial for Earth observation, with improving resolutions driving demand for quantitative evaluation.
• Conventional simulation-based evaluation methods face a speed-accuracy trade-off, limiting real-time applications.
• Existing methods do not fully account for on-board resource constraints or imaging uncertainties.

Purpose of the Study:

• To propose a novel real-time evaluation model architecture for remote sensing satellite clusters.
• To develop an indicator system for assessing satellite cluster effectiveness in observing moving targets.
• To enable rapid, high-quality quantitative evaluations for stakeholders.

Main Methods:

• Established a multi-physical field coupling simulation model for satellite clusters observing moving targets, incorporating resource constraints and imaging uncertainties.
• Developed a moving target observation indicator system and employed correlation analysis for indicator independence.
• Designed and trained a neural network model, optimizing its structure and parameters for rapid evaluation.

Main Results:

• The proposed model architecture facilitates real-time evaluation of remote sensing satellite clusters.
• The indicator system effectively reflects the satellite cluster's orbital performance.
• The optimized neural network achieves high-quality evaluation results in real-time.

Conclusions:

• The developed neural network evaluation model provides a valid and efficient solution for real-time quantitative assessment of remote sensing satellite clusters.
• This approach overcomes the limitations of traditional simulation methods by balancing speed and accuracy.
• The model enhances the practical utility of remote sensing satellite data for Earth observation.