基于轨迹重构与贝叶斯推理的空中机器人灯塔距离测绘技术  

Distance Measurement Technology for Aerial Robot Lighthouses Based on Trajectory Reconstruction and Bayesian Inference

作　　者：覃学宁 钟辉 岳志伟 何瑞冠 冷泉 李飞 QIN Xuening;ZHONG Hui;YUE Zhiwei;HE Ruiguan;LENG Quan;LI Fei(Sansha Navigation Center Ministry of Transport,Haikou 570000,China)

机构地区：[1]交通运输部南海航海保障中心三沙航标处,海口570000

出　　处：《计算机测量与控制》2024年第10期291-297,共7页Computer Measurement &Control

摘　　要：灯塔在目前仍然发挥着重大作用,可以在导航和定位过程中提供重要的参考信息;传统方法通常采用无人机导航,但受多路径以及动态环境的干扰,导航误差较大;为精准灯塔的工作范围,提出基于轨迹重构与贝叶斯推理的空中机器人灯塔距离测绘技术;采用捷联式惯性导航重构惯性轨迹,对空中机器人惯性导航;采集灯塔灯光距离测绘数据,构建基于贝叶斯推理的多线索视觉注意模型,获取灯塔距离显著图,搭建航拍摄像机成像模型,计算灯塔灯光照射距离;测试结果表明,该技术平均导航误差最终仅为0.132 mm,且测绘时间较短,通过多次测试后平均灯塔灯光距离测绘误差仅为27.41 cm,整体测绘误差小。Lighthouses still play an important role today,providing important reference information during navigation and positioning.The traditional method usually uses UAV navigation,but due to the interference of multi-path and dynamic environment,the navigation error is large.An aerial robot lighthouse distance mapping technique based on trajectory reconstruction and Bayesian inference is proposed for accurate lighthouse working range.The inertial trajectory is reconstructed by strapdown inertial navigation to inertial navigation of aerial robot.The light distance mapping data of the lighthouse was collected,a multi-cue visual attention model based on Bayesian inference was constructed,the significant map of the lighthouse distance was obtained,the imaging model of the aerial camera was built,and the illumination distance of the lighthouse light was calculated.The test results show that the average navigation error of this technology is only 0.132 mm,and the mapping time is short.After multiple tests,the average mapping error of lighthouse light distance is only 27.41 cm,and the overall mapping error is small.

关 键 词：轨迹重构 惯性导航 贝叶斯推理 空中机器人 灯塔距离测绘 

分 类 号：TU19[建筑科学—建筑理论]