超参数自适应的MOEA/D-DE算法在翼型气动隐身优化中的应用

MOEA/D-DE 算法易于实现，被广泛应用于处理多目标优化问题，但其超参数CR 和F 对算法性能影响较大。基于MOEA/D-DE 算法框架、利用Sobol 全局灵敏性分析方法对差分进化算子中的交叉控制参数CR进行改进，使用莱维飞行策略控制比例因子F，使算法中的超参数拥有自适应能力，得到超参数自适应的MOEA/D-DE 算法——MOEA/D-DEAH 算法；对MOEA/D-DEAH 算法、不同超参数设置的MOEA/D-DE算法和NSGAII 算法进行函数测试和翼型气动隐身优化算例对比。结果表明：MOEA/D-DEAH 算法性能良好，具有较强的鲁棒性，气动隐身优化效果也比其他算法更好。     

Low Earth orbit constellation design using a multi-objective genetic algorithm for GNSS reflectometry missions

Spaceborne global navigation satellite system reflectometry (GNSS-R) is an innovative bistatic radar remote sensing technique utilizing low Earth orbit (LEO) based GNSS-R instruments to acquire GNSS L-band opportunistic signals for measuring geophysical parameters. A GNSS-R LEO constellation with an optimization design for its specialized missions is very significant and necessary. However, the constellation design involves multi-parameter and multi-objective optimization, and the classical analytic solution is not capable of such a complicated issue. This study proposes a multi-objective LEO constellation design method with a genetic algorithm (GA) and presents a framework for designing two GNSS-R LEO constellations, termed “lower-latitude constellation” for typhoons and hurricanes observation in the tropics and “global constellation” for global geophysical parameter measurements. Then, the observation capability of both designed constellations is evaluated in terms of the number of reflection points, spatial coverage density, and revisit time to verify the GA efficiency in LEO constellation design. Results show that the two designed LEO constellations with high fitness function values possess optimal orbit parameter set configuration and outperform the existing CyGNSS constellations in observation performance. Compared with CyGNSS, the number of reflection points observed by the lower-latitude constellation and the global constellation increases by 38% and 45%, as well as the spatial coverage density increases by 28% and 36%. The revisit time for the lower-latitude constellation is reduced by 0.29 h, whereas the revisit time for the global constellation increases by one hour.     

Reinforcement learning based dynamic distributed routing scheme for mega LEO satellite networks

Recently, mega Low Earth Orbit (LEO) Satellite Network (LSN) systems have gained more and more attention due to low latency, broadband communications and global coverage for ground users. One of the primary challenges for LSN systems with inter-satellite links is the routing strategy calculation and maintenance, due to LSN constellation scale and dynamic network topology feature. In order to seek an efficient routing strategy, a Q-learning-based dynamic distributed Routing scheme for LSNs (QRLSN) is proposed in this paper. To achieve low end-to-end delay and low network traffic overhead load in LSNs, QRLSN adopts a multi-objective optimization method to find the optimal next hop for forwarding data packets. Experimental results demonstrate that the proposed scheme can effectively discover the initial routing strategy and provide long-term Quality of Service (QoS) optimization during the routing maintenance process. In addition, comparison results demonstrate that QRLSN is superior to the virtual-topology-based shortest path routing algorithm.