基于组件匹配的变电站快速建模方法

RAPID MODELING METHOD FOR SUBSTATIONS BASED ON COMPONENT MATCHING

针对智能变电站模型重建问题,提出一种基于组件匹配的快速建模方法。在数据预处理阶段,变电站激光扫描数据中的电力线及地面数据将被自动去除,各独立部件也会被自动分割出来。在此基础上,算法自动探测变电站内的重复部件。对每一种重复部件,算法自动检索模板库以获取最相似的模型并同时将模型置入点云模型中相对应的位置以完成重建。与传统的人工建模方式相比,算法结合变电站中模型的重复性及标准性特点,将基于组件匹配的建模思想引进到变电站建模过程中,这一举措显著地提高了建模效率和精度。特别是针对大规模的变电站模型,算法的重建效率更为明显。大量的实验结果验证了基于组件匹配的快速建模方法的有效性及鲁棒性。

This paper presented a rapid modeling method based on component matching for the problem of smart substation model reconstruction. In a preprocessing step,algorithm automatically removed all power lines and the ground point cloud followed by segmenting out all individual components. On this basis,we explored the repeatability and regularity of man-made structures in substations,which was the crux of automatic reconstruction. Then,for each kind of repetition,we performed component matching to achieve the final reconstruction. The 3D component was automatically chosen and located by proposed template assembly algorithm which balanced the model matching accuracy and structural regularity. In contrast to traditional manual modeling method,algorithm employed the idea of component matching by taking advantage of the repeatability and regularity of substation models which significantly enhanced the efficiency and accuracy,especially in the reconstruction of large-scale power substations. A variety of substation scenes have been tested to verify the effectiveness and the robustness of the proposed method.