基于遗传算法的多星座选星方法  

Satellite Selection with Genetic Algorithm Under Multi-Constellation

作　　者：宋丹[1] 许承东[1] 胡春生[1] 张鹏飞[1] 

机构地区：[1]北京理工大学宇航学院,北京100081

出　　处：《宇航学报》2015年第3期300-308,共9页Journal of Astronautics

基　　金：国家自然科学基金(61173077);国家863计划(2011AA120505)

摘　　要：提出了一种基于遗传算法的选星方法,该方法能够准确、快速地实现多个全球导航卫星系统(GNSS)下选星颗数大于4的选星。经过大量试验确定了遗传算法选星的选择、交叉、变异等关键参数,并通过在交叉和变异中穿插染色体优选过程克服了遗传算法的选星结果容易陷入局部最优的缺点。以不同GNSS星座个数和不同选星颗数下的选星仿真校验了所设定遗传参数的有效性和可靠性。仿真结果表明,利用遗传算法可以高效、准确地实现选星,该方法特别适用于多星座多选星颗数的情况。Satellite selection with genetic algorithm（ GA） is proposed in this paper. With this method,selecting more than four satellites under multiple global navigation satellite systems（ GNSS） can be achieved accurately and rapidly. By inserting the chromosome merit-based selection in the process of crossover and mutation,the local optimum can be avoided substantially when selecting satellites with GA. The GA parameters for satellite selection are chosen after a series of simulation experiments. Then the performance of satellite selection with GA is confirmed to be remarkable by use of the simulation experiments under different numbers of GNSS constellations and selected satellites. The selected satellites are optimum and the calculation time of satellite selection is acceptable.

关 键 词：选星 遗传算法 几何精度衰减因子 多GNSS星座 

分 类 号：P228.1[天文地球—大地测量学与测量工程]