Star Identification Algorithms And The Aerocraft Attitude Determination Research

As a kind of vector attitude sensor, the star sensors use vector observations from the star to measure and determine the attitude, and gain popularity for their high accuracy, without attitude accumulation errors and other advantages. During the application process, two main problems must be solved: the first is the star identification, and the second is the attitude determination based on the sensor.This paper is focusing on the star identification algorithm, as well as the attitude determination based on the vector observations by using the star sensor. So the topics presented and discussed in this paper are as follows:1. Define the reference frames and quaternion theory used in the star identification algorithm and attitude determination. Moreover the frame transformation and the dynamics equations are deducted.2. Several suggestions of choosing navigation stars and building the star database are put forward, according to the statistic rules of the stars in the field of view (FOV). In addition, three star identification algorithms are analyzed and compared.3. Deduce the three attitude determination methods (TRIAD, q-method, QUEST) in terms of the vector observations. Furthermore giving some practical examples and analyze the accuracy and application range of each method by Matlab simulation.4. Besides the traditional EKF, this paper presents a new method UKF, and uses it in the attitude estimation. Comparing the two simulation results, we can see whether the performance is batter than EKF.5. Finally, we built a simulation platform to realize each step according to the real conditions, according to the star sensor's application process, which includes choosing navigation stars, the random direction test in FOV, and the attitude determination.