Research On Optimization Algorithm Of Ground Observation Station For Navigation Satellite Precise Orbit Determination

Ground observation station optimization is a key link in process of satellite orbit determination.Inappropriate selection will directly affect accuracy of satellite orbit determination.With rapid development of satellite orbit determination technology,Global Navigation Satellite System(GNSS)observation station optimization has higher requirements.How to quickly and effectively optimize global GNSS observation stations is a key issue that needs to be solved urgently to improve accuracy and efficiency of satellite orbit determination.This paper takes International GNSS Service(IGS)observation stations as research object,and focuses on observation station post and real-time optimization research of Global Positioning System(GPS)and BeiDou Navigation Satellite System(BDS).It’s carried out from four aspects:evaluation index of observation station optimization,improved genetic algorithm for post optimization,improved Delaunay triangulation for real-time optimization and design of observation station optimization system.It aims to select the best combination,and improve accuracy and efficiency of satellite orbit determination.The main research contents of the paper are as follows:(1)Improve evaluation index of observation station optimization to evaluate observation quality and geometric configuration.For observation station post optimization,five single indexes,including data integrity rate,multipath error,derivative of ionospheric delay,observation station coordinate variance and signal-to-noise ratio,are used to evaluate observation quality.And through entropy calculation to comprehensively process observation quality to obtain weight matrix of observation stations,and then construct the Observation Quality Weighted Dilution of Precision(OQDOP)evaluation index used for post optimization.For observation station real-time optimization,data integrity rate,minimum number of satellite observations,minimum number of epoch satellites,and dual-frequency condition are used as observation quality real-time evaluation indexes for observation station preliminary screening,and the minimum distance sum evaluation index is generated through Delaunay triangulation.The experimental results show for post optimization evaluation index,the orbit determination accuracy increases with the decrease of OQDOP.For real-time optimization evaluation index,the greater the minimum distance sum,the higher the orbit determination accuracy.(2)Aiming at traditional genetic algorithm problems of decreasing population diversity in early stage and crossing and mutating excellent individuals in late stage,an improved genetic algorithm is proposed.It uses combination of elite retention strategy and probabilistic survival strategy for population selection.Crossover rate and mutation rate decay linearly with evolutional times,so that the population diversity is increased at early evolution,and crossover rate of excellent individuals is reduced at late evolution.Meanwhile,information entropy and variance are introduced to judge algorithm iterative convergence more accurately,so that the algorithm performance can be exerted better.Experimental results show for post optimization of GPS observation stations,improved genetic algorithm improves one-dimensional(1D)orbit determination accuracy by 13.55%and 20.16%,and optimization efficiency by 17.58%and 31.85%,respectively,compared with traditional genetic algorithm and grid method.For BDS post optimization,1D orbit determination accuracy of MEO satellites is improved by 12.70%and 23.78%,IGSO satellites is improved by 10.06%and 17.10%,and optimization efficiency is improved by 15.37%and 82.21%,respectively.(3)Aiming at problems of inaccurate uniformity measurement and emphasis on geometric configuration and dilute observation quality of traditional Delaunay triangulation optimization algorithm in observation station real-time optimization,an improved Delaunay triangulation optimization algorithm is proposed.The algorithm preliminarily screens observation stations based on real-time observation quality indexes,and generates observation station combination set through grid method.Finally,the optimal observation stations combination is obtained according to the principle of maximizing the minimum distance sum.Experimental results show compared with traditional optimization algorithm,improved Delaunay triangulation optimization algorithm improves 1D orbit determination accuracy of GPS satellites by 15.09%,and BDS MEO and IGSO satellites by 15.78%and 14.29%,respectively.The optimization efficiency is equivalent to traditional optimization algorithm.(4)To meet automation and visualization requirements of satellite orbit determination observation station optimization,GNSS observation station real-time optimization system is designed and implemented.The system is divided into login module and optimization module.It can perform observation station optimization of GPS and BDS,and has functions such as observation station number input,observation quality control,optimization progress display and result distribution display.The interface is simple,fully functional,and highly integrated.The test results show the optimization system has perfect functions,steady performance,good portability and reusability.