The optimization of low Earth orbit satellite constellation visibility with genetic algorithm for improved navigation...

Area of Science:

• Satellite Constellation Design
• Optimization Algorithms
• Orbital Mechanics

Background:

• Low Earth Orbit (LEO) satellite constellations face visibility optimization challenges.
• Ensuring navigation accuracy and dynamic robustness is critical for LEO constellations.
• J2 perturbation effects impact long-term constellation performance.

Purpose of the Study:

• To propose an enhanced framework for LEO satellite constellation visibility optimization.
• To improve navigation accuracy and dynamic robustness of satellite constellations.
• To provide a technical foundation for next-generation Global Navigation Satellite System (GNSS) enhancements.

Main Methods:

• Developed an adaptive parallel Genetic Algorithm (GA) framework.
• Designed a hybrid constellation integrating polar, Walker, and Orthogonal Circular Orbits.
• Incorporated a dynamic relaxation factor (γ) to compensate for J2 perturbations.
• Implemented adaptive parameter adjustment using population diversity entropy.
• Utilized a parallel fitness evaluation strategy for multi-core architectures.
• Employed a simplified fitness function design.

Main Results:

• Achieved an average of 14.3 visible satellites in a 100-satellite scenario, outperforming D-NSDE by 3.6%.
• Reduced Position Dilution of Precision (PDOP) to 2.3 with 95.6% global coverage.
• Demonstrated high robustness with only a 3.5% coverage drop after single satellite failure.
• Maintained 94.8% coverage after 10 years of orbital perturbation.
• Achieved PDOP ≤ 2.8 with convergence times under 210s for 500-1000 satellite scenarios.
• Outperformed Particle Swarm Optimization and Dynamic Non-dominated Sorting Differential Evolution in visibility, robustness, and efficiency.

Conclusions:

• The proposed adaptive parallel GA framework effectively optimizes LEO satellite constellation visibility.
• The framework offers superior navigation accuracy, dynamic robustness, and computational efficiency.
• This research provides a significant advancement for future GNSS development.