УСУНЕННЯ АНОМАЛЬНИХ ПОХИБОК ПОЛЬОТНОГО ГЕОМЕТРИЧНОГО КАЛІБРУВАННЯ

• Datum: 2025
• 1. Verfasser: Tkachenko , A.I.
• Format: Artikel
• Sprache: English
• Veröffentlicht: V.M. Glushkov Institute of Cybernetics of NAS of Ukraine 2025
• Schlagworte: польотне геометричне калібрування; космічний апарат; знімальна камера; зоряний датчик; наземні маркери; розмитий спостережник; цикл; збіжність оцінок
• Online Zugang: https://jais.net.ua/index.php/files/article/view/697
• Назва журналу: Problems of Control and Informatics

An in-flight geometric calibration (further — calibration) is considered here as procedure of refining mutual attitude parameters of the onboard imaging camera and star tracker. The problem of calibration is solved with using of observations of geo-referenced landmarks from the orbit. Usually above-mentioned procedure is preceded by preflight calibration which is performed in lab or in industrial conditions with proper equipment and is relatively complicated technologically. A necessity of in-flight geometric calibration takes place for instance when preflight calibration does not ensure acceptable accuracy of ground objects georeferencing by means of space snapshots received with use of optical-electronic complex, or if indefiniteness of camera’s angular attitude relatively to star tracker accumulates in a process of exploiting of the spacecraft on the orbit. Obviously in common case it is necessary to take into account a possibility of appearance of inadmssibly big errors in accessible estimations of camera and star tracker mutual attitude parameters. Usually construction of measuring equations is accompanied with linearization of them. Omitted non-linear effects may strictly limit attainable calibration accuracy out of convergence domain for estimations if highly big initial errors take place. In such situations it is desirable to apply a method of revealing and exclusion of unacceptably big calibraton errors. A version of methods to be applied with such purpose is developed in this work. They are based on two effects: high convergence characteristics of estimation algorithm — fuzzy state observer — and succession of calculations in which each measuring equation, instead of invariable initial angular error, takes into account and estimates its remainder after processing of previous measurings. This the second effect diminishes the omitted nonlinear component of the error and in such a way improves the convergence of estimates. After processing of all accessible measurings and correction of the searched parameters, the cycles of processing of the same measurings are repeated with use of corrected parameters. On the basis of comparison of previous and following cycles, the conclusion about the level of initial error and convergence of the estimation is made.