文档介绍：中国科技论文在线
基于遗传算法与自适应算法的位姿估测与
滤波#
冷青凡，薄煜明，赵高鹏，刘鲁江
5 （南京理工大学自动化学院，南京 210094）
摘要：本文研究了双目视觉下非合作目标交会对接平移位姿的估测。在坐标系建立与系统建
模的基础上，设计了遗传 TRIAD 算法进行平移位姿的初步估测，将非支配排序遗传算法与
矢量处理改进下 TRIAD 算法相结合，能综合而灵活地提高精确度，并保证鲁棒性与稳定
性。设计了系统噪声协方差自适应估计的卡尔曼滤波器对初步估测结果进行滤波，针对系
10 统噪声协方差估计设计了改进型自适应指数法，进一步提高了估测的精度与鲁棒性。仿真
证明了算法设计的有效性与优越性，适合非合作目标交会对接系统的工程应用。
关键词：工程控制论；非支配排序遗传算法；TRIAD 算法；自适应；滤波
中图分类号：TP301
15
The estimation and filtering of position and pose based on
20 genetic algorithm and adaptive algorithm
leng qing fan, Bo Yuming, Zhao Gaopeng, Liu Lujiang
(School of automation Nanjing University of Science and Technology，Nanjing 210094)
Abstract: In this paper, the position and pose of the non cooperative target rendezvous and
25 docking under the binocular vision is studied and estimated. The genetic TRIAD algorithm is
designed for a preliminary estimation of the translational pose basing on the modeling of
coordinate systems and engineering systems, combining non-dominated sorting genetic algorithm
and TRIAD algorithm under improved vector processing. It can improve accuracy synthetically as
well as neatly, and ensure the robustness and stability. A new Kalman filter is designed for
30 estimating adaptively the covariance of the system noise, in which a modified method of adaptive
index is projected, further improving the accuracy and robustness of the