Sun Quansheng, Liao Tianjun, Du Haibo, Zhao Yinfeng, Chen Chih-Chiang
School of Electrical Engineering and Automation, Hefei University of Technology, Hefei 230009, China.
Academy of Military Sciences, Beijing 100850, China.
Sensors (Basel). 2023 Jan 1;23(1):447. doi: 10.3390/s23010447.
The merging of environmental maps constructed by individual UAVs alone and the sharing of information are key to improving the efficiency of distributed multi-UAVexploration. This paper investigates the raster map-merging problem in the absence of a common reference coordinate system and the relative position information of UAVs, and proposes a raster map-merging method with a directed crossover multidimensional perturbation variational genetic algorithm (DCPGA). The algorithm uses an optimization function reflecting the degree of dissimilarity between the overlapping regions of two raster maps as the fitness function, with each possible rotation translation transformation corresponding to a chromosome, and the binary encoding of the coordinates as the gene string. The experimental results show that the algorithm could converge quickly and had a strong global search capability to search for the optimal overlap area of the two raster maps, thus achieving map merging.
仅由单个无人机构建的环境地图的合并以及信息共享是提高分布式多无人机探索效率的关键。本文研究了在没有公共参考坐标系和无人机相对位置信息的情况下的栅格地图合并问题，并提出了一种基于定向交叉多维扰动变分遗传算法（DCPGA）的栅格地图合并方法。该算法使用反映两个栅格地图重叠区域差异程度的优化函数作为适应度函数，每个可能的旋转变换对应一个染色体，坐标的二进制编码作为基因串。实验结果表明，该算法能够快速收敛，具有很强的全局搜索能力，能够搜索到两个栅格地图的最优重叠区域，从而实现地图合并。
