ALGORITHM FOR TRAJECTORY INFORMATION PROCESSING OF RADAR MEASURING COMPLEXES BASED ON K-MEANS CLUSTERING
JSC «Kronshtadt technologies», N.S. Kondibaev, N.A. Kupriyanov, S.Z. Kurakin · H&ES Research · 2020
The problem situation caused by the need to improve the accuracy of determining the location of observed space objects by radar measuring complexes is considered. It is shown that the accuracy characteristics of radar measuring systems depend on the contribution of various measurement errors, the most significant of which for modern highly informative observation tools are atmospheric, determined by helio-geophysical conditions of operation. It is noted that the currently applied approaches to compensation of atmospheric measurement errors do not allow taking into account the heliogeophysical conditions of operation of radar measuring complexes with high time and spatial resolution. It is shown that when processing coordinate information, sporadic changes in heliogeophysical operating conditions are not fully taken into account, which leads to an increase in errors in measuring the location of space objects and a decrease in the information capabilities of radar measuring complexes. The idea of using the discrepancy of measurements by the radar measuring complex of the location of space objects for which there is a priori coordinate information is considered. The results of combining groups of space object location measurements using cluster analysis methods are shown. The results of computer modeling are presented and it is shown that the proposed approach allows us to determine the areas where the influence of heliogeophysical conditions increases the measurement errors. The main calculation relations are described and the stages of the algorithm for trajectory processing of information from radar measuring complexes based on clustering by the K-MEANS method are shown. Further directions of using the results of the algorithm are proposed, which con sist in using data on formed clusters for processing information about space objects for which there is no a priori coordinate information. It is shown that the proposed approach can be applied in order to improve the information capabilities of radar measuring systems.