基于自适应格网数字水深模型的水下地形匹配定位算法

Underwater terrain matching algorithm based on adaptive grid digital depth model

为了解决当前的水下地形匹配定位算法未考虑格网大小随海底地形变化而自动调整的问题,提出了一种基于自适应格网数字水深模型的水下地形匹配定位算法。首先,引入基于四叉树的自适应格网模型,改进其局部格网的相似性评估指标与构网约束方法;然后,设计匹配区的确定及待匹配航迹的选取策略,给出待匹配航迹的水深值的计算方法,构建出目标匹配定位的地形相关组合算子。实验结果表明:提出的算法相比于基于等距离的规则格网模型的算法有以下优势:1)在地形特征丰富区域的定位精度明显提高;2)可以避免出现地形特征越明显匹配定位精度越低的情况;3)能一定程度地克服水深系统误差对定位精度的影响。

The underwater terrain aided matching algorithm does not take into account that the grid space of reference map should automatically be adjusted with seabed terrains. To solve this problem, a new terrain matching algorithm based on adaptive Digital Grid Model is presented. First, an adaptive grid model based on quadtree is put forward to improve its local grid evaluation index and the construction method. Then, the strategy used to select the matching regions is designed, and the algorithm for calculating the track depths ready to be matched is given. Finally, the topographic correlation combination operator for object matching and localization is established. Experiment results show that the proposed method is superior to the traditional algorithm based on regular Digital Depth Model in that： 1） the new method can capture more precious location in the areas with substantial features; 2） the new method can avoid the situation that more obvious terrain features lead to more lower location precision; and 3） the new method can significantly overcome the system error＆#39;s influence on the location precision.