基于点云匹配的智能工厂室内定位算法研究

Research on Indoor Positioning Algorithm for Intelligent Factory Based on Point Cloud Matching

针对智能工厂中的无人叉车系统需要精准的室内定位来满足装卸货工序的需求,研究了基于SLAM的定位算法以及基于反光柱的两种激光点云数据匹配定位算法,分析了这两种方法的优势和问题,并提出了一种新的融合定位算法。算法以高精度的反光柱定位算法为核心,通过SLAM定位来弥补反光柱数量不足时的系统延迟和精度损失问题,从而实现了一套连续、稳定、可靠的室内定位系统,其定位精度可以满足无人叉车日常工作的需求,也可以直接用于具体的工业生产流程中。

In this paper,we implemented a SLAM-based positioning algorithm and a reflective-pillar-based positioning algorithm to meet the demands of accurate indoor positioning of unmanned forklift systems in smart factories.We analyzed the advantages and problems of these two methods,and then proposed a new fusion-based positioning algorithm.The algorithm mainly uses the high-precision reflective-pillar positioning algorithm,and compensates the system delay and accuracy loss with the SLAM positioning algorithm,thus realizes a continuous,stable and reliable indoor positioning system,which can meet the needs of unmanned forklifts in industrial production environments.