基于UKF滤波加权C-T融合算法的无人驾驶单轨吊双标签UWB定位  

UWB localization of unmanned monorail crane with dual tags based on UKF weighted C-T fusion algorithm

作　　者：吕玉寒 张牧野 鲍久圣[1] 杨阳[2] 杨健健[3] 王茂森 LYU Yuhan;ZHANG Muye;BAO Jiusheng;YANG Yang;YANG Jianjian;WANG Maosen(School of Mechatronic Engineering,China University of Mining and Technology,Xuzhou 221116,China;Shijiazhuang Coal Mining Machinery Co.,Ltd.,Shijiazhuang 051431,China;School of Mechanical and Electrical Engineering,China University of Mining and Technology-Beijing,Beijing 100083,China)

机构地区：[1]中国矿业大学机电工程学院,江苏徐州221116 [2]石家庄煤矿机械有限责任公司,河北石家庄051431 [3]中国矿业大学(北京)机械与电气工程学院,北京100083

出　　处：《煤炭科学技术》2024年第S2期221-235,共15页Coal Science and Technology

基　　金：国家煤矿安全监察局科研项目(2019-行管司-022-02);江苏高校优势学科建设工程资助项目(PAPD);中国高校产学研创新基金资助项目(2024HT012)。

摘　　要：单轨吊机车是煤矿井下重要的辅助运输设备之一,目前正在朝着智能化和无人化的方向发展。为了提高无人驾驶单轨吊定位精度,开展了基于无迹卡尔曼滤波(UKF)加权C-T融合算法的双标签超宽带(UWB)定位方法研究。首先,根据单轨吊车身的结构特征,设计了包含双标签定位信息采集层、定位数据传递层和定位坐标解析层的双标签UWB定位系统;其次,将UWB中Chan算法的定位结果作为Taylor算法的初始值,保障了Taylor算法的收敛性和计算效率;再次,通过预设的单轨吊车身长度与双标签定位数据得到定位补偿error,将error代入Taylor算法以进一步提高定位精度,仿真结果表明优化后算法的定位精度提高了44%;然后,使用UKF对加权C-T融合算法进行滤波优化,提高了定位系统在非视距(NLOS)环境中的定位精度,仿真结果表明UKF滤波优化后的定位精度在直行路段提高了7.8%以上,在弯道路段中提高了10.6%以上,且随着NLOS误差的增大,定位效果明显提升;最后,在石煤机试验场进行单轨吊实车试验,结果表明:基于UKF滤波的双标签加权C-T融合定位算法使单轨吊静态定位精度小于20cm,动态定位精度小于30cm,整体定位精度达到分米级,稳定性和可靠性也得到提高,可满足单轨吊井下无人驾驶定位需求。研究分米级精度的单轨吊定位系统是矿井单轨吊实现智能化、无人化高效运输的重要保障。Monorail cranes are crucial auxiliary transportation equipment in underground coal mines,currently advancing towards intelligent and unmanned operation.To enhance the precision of unmanned monorail crane positioning,research has been conducted on a dualtag UWB positioning method based on the UKF filtering and weighted C-T fusion algorithm.Firstly,considering the structural characteristics of monorail cranes,a dual-tag UWB positioning system was designed,comprising a dual-tag positioning information collection layer,positioning data transmission layer,and positioning coordinate resolution layer.Secondly,the Chan algorithm's UWB positioning results were employed as initial values for the Taylor algorithm,ensuring the convergence and com-putational efficiency of the Taylor algorithm.Additionally,by predefining the monorail crane length and obtaining positioning compensation error from dual-tag positioning data,the error was incorporated into the Taylor algorithm to further enhance positioning accuracy.Simulation results demonstrated a 44% improvement in positioning accuracy with the optimized algorithm.Subsequently,the Unscented Kalman Filter(UKF)was applied for filtering optimization of the weighted C-T fusion algorithm,enhancing the positioning system's accuracy in Non-Line-of-Sight(NLOS)environments.Simulation results indicated that the UKF-filtered opti-mization increased positioning accuracy by over 7.8% on straight segments and over 10.6% on curved segments.Moreover,as NLOS errors increased,positioning effectiveness significantly improved.Finally,real-world experiments were conducted in a coal mine monorail test field,revealing that the dual-tag weighted C-T fusion positioning algorithm based on UKF filtering achieved static positioning accuracy below 20 cm,dynamic positioning accuracy below 30 cm,and an overall positioning accuracy at the decimeter level.Stability and reliability were also enhanced,meeting the positioning requirements for unmanned monorail cranes in under-ground environments.The research on d

关 键 词：单轨吊 超宽带(UWB) 加权C-T融合算法 双标签 无迹卡尔曼滤波(UKF) 

分 类 号：TP52[自动化与计算机技术]