针对VINS-Fusion在无人机同步定位与建图(Simultaneous Localization And Mapping,SLAM)回环检测中由词袋模型所产生的算力消耗较大以及检测表现不佳等问题,文章在VINS-Fusion框架中提出了基于卷积编码器的回环检测改进算法。图像通过卷积编码器得到类方向梯度直方图序列值,通过该值来进行图像的检测匹配。在KITTI数据集上的实验运行表明,该算法在减少了回环检测时间的同时,提高了整个无人机SLAM系统的鲁棒性。
Electricity being a basic national infrastructure for a country, its security, reliability and quality are the most important parameters for the network managers. Several methods are generally used to improve the voltage quality more and more. However, most of the means implemented depend on external factors independent of the network managers or require huge regular financial resources. The method used in this paper is the loopback, applied to the Southern Interconnected Grid (SIG) of Cameroon, which is the largest network in the country. The procedure used takes into account nodes experiencing huge voltage drops and network constraints. The chosen loopback scenario results in a clear improvement of the voltage plan in this network, and also a discharge of the transformers, a considerable decongestion of the lines, a reduction of the power losses and a significant reduction of the thermal placement used for improvement of the voltage profile.
