Research On Calibration Methods For Star Tracker/Gyro Based Attitude Determination System

Remote sensing images acquired from earth observation are increasingly important in almost all industries.In order to get high-quality remote sensing images,more effort should be paid in improving real-time and off-line satellite attitude determination accuracy.High-accuracy attitude determination systems usually consist of star sensors and gyroscopes.Theoretically,attitude information provided by star sensor is higher than any other attitude sensors.However,due to the vibration,electromagnetic influence,gravity,cosmic radiation and complex working environment,various error exists in the star sensor/gyro based system.To achieve higher attitude determination results,calibration techniques are employed as it is the certain way that we can improve the attitude determination and control accuracy.This paper aims at solving the problem that the accuracy of attitude determination systems is greatly decreased by multiple errors of attitude sensors.Star sensor,gyroscope and attitude determination systems consist of these two kinds of sensors are the research subjects in this paper.Internal parameters,installation parameters that influence the attitude determination system is modeled and estimated.The main content is as follows:Firstly,commonly used coordinate systems and transformation method,attitude description method and attitude dynamics are described in detail.Star sensor and gyro models that contains almost all errors in the process of attitude determination are introduced in this chapter.Secondly,optical errors of star sensors are modeled and calibrated.Optical error calibration can significantly improve the datum of the whole attitude determination system,so optical calibration is necessary.Imaging model and optical error are constructed.A machine-learning based method is proposed in this chapter to calibrate lens distortion,focal length error and main point offset.The calibration process is also given and the method is proved by numerical simulation.Furthermore,low frequency error of star sensor is modeled as Fourier series and calibrated.The coupling influence between LFE parameters and gyro drift is discussed in detail.A maximum-likelihood-estimation based bidirectional adaptive filtering method is developed.Meanwhile,the calibration situation that cyclical error exists in gyro drift is discussed specially.Finally,based on the calibration results of Chapter 3 and 4,installation errors calibrationof star sensor/gyro based attitude determination system is presented.Datum determination and calibration of multi-star sensor system is developed,and the dissemination of LFE is discussed particularly.The installation and scale factor of gyroscopes are calibrated in this chapter and the corresponding calibration process of gyroscope is also given.Moreover,the calibration technique is summarized and the overall calibration process is presented,which provide a clear view of the relationship between calibration processes of various errors and give a basic reference calibration order.