论文部分内容阅读
由于光学成像系统的非线性几何畸变,使得星敏感器所获星图与理想星图有一定的差别,为完成高精度定姿,必须对畸变星图进行校正。首先,介绍了光学系统的畸变原理,并建立了以径向几何畸变为主的星图畸变模型。之后引入改进遗传算法对畸变参数进行了优化:采用引进成长算子的二进制编码,避免了算法陷入伪极值;通过改进适应度函数和适时调整变异概率,避免了计算中产生的早熟收敛问题。通过与基本遗传算法的比较结果表明,该方法不仅能够降低75.3%的相对误差,而且还提高了16ms的畸变校正速度,基本能够满足星图识别和姿态确定对精度高、实时性强等性能的要求。
Due to the nonlinear geometric distortion of the optical imaging system, the difference between the star image and the ideal star image obtained by the star sensor is somewhat different. In order to achieve the high-accuracy attitude setting, the distortion star image must be corrected. First of all, the principle of optical system distortion is introduced, and the star distortion model based on radial geometric distortion is established. After that, the improved genetic algorithm is introduced to optimize the distortion parameters. By introducing the binary code of the growth operator, the algorithm avoids the plunge into the pseudo-extreme value. By improving the fitness function and adjusting the mutation probability in time, premature convergence problem in computation can be avoided. The comparison with the basic genetic algorithm shows that this method can not only reduce the relative error of 75.3%, but also improve the distortion correction speed of 16ms, which can basically meet the requirements of high resolution and real-time performance of the satellite image recognition and attitude determination Claim.