基于遗传算法的相对位姿确定方法

Relative Pose Determination Method Based on Genetic Algorithm

针对空间非合作航天器之间接近操作过程中的相对位姿单目视觉确定问题,在充分利用目标航天器本身的特征结构信息的基础上,提出一种基于遗传算法的相对位姿求解方法。该方法结合逆投影原理构建的优化模型,将相对位姿的求解问题转化为非线性优化问题,再将其目标函数转化为只含有旋转角度的目标函数,然后采用对非线性优化问题求解优良的遗传算法进行解算,得到旋转角度进而求得旋转矩阵。最后对该算法进行数学仿真,仿真结果表明该方法的有效性和可靠性。

Aiming at the problem of relative pose determination using monocular vision in the near-space operation of space non-cooperative spacecraft, a method of relative pose solving using genetic algorithm is proposed by making full use of the feature structure information of target spacecraft. The method combines the inverse projection principle to construct an optimization model, and transforms the problem of relative pose into a nonlinear optimization problem. Then, the objective function is transformed into an objective function with only rotation angle. After that, the nonlinear genetic algorithm is used to solve the nonlinear optimization problem, and the rotation angle is obtained,and then the rotation matrix is obtained.Finally, the algorithm is mathematically simulated, and the simulation results show the validity and reliability of the proposed method.