| High-precision attitude determination is the prerequisite for high-precision attitude control.The star sensor is the highest accuracy attitude sensor and be used commonly.In the high dynamic case,the star sensor will be out the problem of star point trailing and blurring,the traditional star point extraction algorithm is too low precision.In addition,in order to meet the needs of satellite high-precision attitude determination commonly used star sensor/gyroscope integrated attitude determination.For those questions,the main works of this paper are as follows:Firstly,introduce the basic principle of star sensor attitude determination.Design the rules for selecting navigation stars and a method for star simulation.The common star recognition methods and attitude calculation methods are summarized.Secondly,the adaptive trailing star extraction algorithm based on known angular velocity is proposed.The trailing source(relative motion)for high dynamic map is analyzed.Linear model is established and three-dimensional model respectively.Then discuss the different effects on map trailing from different axis rotation.After that puts forward the algorithm of this chapter.Made an important emphasis on the gain coefficient and pointed out that the gain factor should be changed by different angular acceleration.Algorithm simulation shows that the gaussian radius of star distribution model and angular velocity output frequency are two important factors affected the accuracy of the algorithm.When other condition is constant,the greater the gaussian radius and higher frequency of velocity output,the extraction precision is higher.Finally,on the time of the algorithm performance is analyzed.The minimum output frequency is 1 Hz,can meet the requirements of the attitude output frequency of star sensor.Thirdly,design star sensor star pattern recognition and attitude determination algorithm.The star recognition grid algorithm based on three libraries and star stably tracking algorithm are designed.The Euler-q method of attitude calculation and the mutual transformation between the coordinate system of attitude calculation are introduced.Then,using Stellarium software to generate simulated star-picture,the simulation experiment is carried out to verify the validity of the star recognition grid algorithm based on three libraries and reliability of the star stably tracking algorithm.The field experiment of the star sensor prototype is carried out to verify the validity and reliability of the star sensor overall design algorithm.Finally,the star-tracker/gyro integrated attitude determination algorithm based on error quaternion is studied.Extended Kalman Filter algorithm based on error quaternion is established.The establishment process of the filtering algorithm is given,and the state linear equation,the measuring equation and the Kalman filter equation are studied.For the measurement equation,the quaternion observation model and Euler angle observation model which are commonly used in engineering are established respectively.The validity and accuracy of the integrated attitude determination algorithm are verified by simulation.Simulation shows that the accuracy of star sensor measurement,the attitude output frequency and the gyro accuracy on the attitude determination accuracy would greatly affect the attitude accuracy.And the filtering effect of the integrated attitude determination on the star tracker with low precision is more obvious.The research of this paper has certain reference value for improving the accuracy of star tracker and satellite attitude determination. |
PDF Full Text Request