组合卫星导航系统的快速选星方法

分析了选星数目与几何精度因子(GDOP, Geometry Dilution of Precision)及导航运算量的关系,基于遗传算法提出了一种以满足用户定位精度需求为条件的快速选星方法——快速遗传选星法.根据用户需求确定选星数目初值、选星数目最大值和GDOP阈值,构造选星方案的初始种群,在进化代数上限为1的条件下对种群进行选择、交叉和变异运算,获得初始选星解,根据初始解的GDOP与阈值的关系确定是否依据GDOP最小原则对初始解进行优化,直至满足算法终止条件,输出选星解.仿真结果表明,该算法可以在一次进化之内以不低于92.45%的概率满足GDOP阈值在2.5~6的要求,同时可有效降低54.75%以上的导航运算量. 

Fast satellite selection method for integrated navigation systems

On the basis of the relationship between the number of the selected satellites and geometry dilution of precision (GDOP) and navigation computation, a genetic algorithm (GA)-based fast satellite selection method named fast genetic satellite selection method (FGSSM) was proposed. In this method, firstly the initial number of selecting satellites and GDOP threshold was decided according to the precision of user's requirement. And then the initial population of satellite selection solutions was constructed. Under the condition that evolution algebra limit is less than 1, the initial solution was obtained by selection, crossover and mutation operation. Whether the initial solution was optimized based on the principles of GDOP minimum was determined by the relationship between the GDOP of the initial solution and the GDOP threshold. The optimization terminates until the termination condition was satisfied and final solution was obtained. The experiment results show that, the method can satisfy GDOP of the solution at the request of 2.5 to 6 within an evolutionary, with no less than 92.45% probability, at the same time it can effectively reduce more than 54.75% of the navigation computation.