基于多编码器与惯导融合的无轨胶轮车定位方法  

A positioning method for underground trackless rubber wheel vehicle based on fusion of multiple encoders and strapdown inertial navigation system

作　　者：江帆[1,2,3] 康静静 皇行涛 朱真才 周公博[1,2] 彭玉兴 JIANG Fan;KANG Jingjing;HUANG Xingtao;ZHU Zhencai;ZHOU Gongbo;PENG Yuxing(School of Mechatronic Engineering,China University of Mining and Technology,Xuzhou 221116,China;National Key Laboratory of Intelligent Mining Equipment Technology,Xuzhou 221116,China;Postdoctoral Scientific Research Workstation of Xuzhou Mining Group Co.,Ltd.,Xuzhou 221018,China;Zhengzhou Hengda Intelligent Control Technology Co.,Ltd.,Zhengzhou 450016,China)

机构地区：[1]中国矿业大学机电工程学院,江苏徐州221116 [2]智能采矿装备技术全国重点实验室,江苏徐州221116 [3]徐州矿务集团有限公司博士后科研工作站,江苏徐州221018 [4]郑州恒达智控科技股份有限公司,河南郑州450016

出　　处：《煤炭科学技术》2024年第S2期284-293,共10页Coal Science and Technology

基　　金：国家自然科学基金面上资助项目(52374163);江苏省自然科学基金杰出青年基金资助项目(BK20230047);中国博士后科学基金面上资助项目(2023M732966)。

摘　　要：针对煤矿井下传统无轨胶轮车定位定向技术精度低、可靠性差等问题,提出了一种基于多编码器与惯导融合的无轨胶轮车定位方法。根据多编码器数据特性,建立了无轨胶轮车轮式里程计运动学模型,进而推导了基于多编码器的航位推算算法。针对上述航位推算算法单独依赖编码器数据易导致定位发散的问题,将轮式里程计与捷联惯导姿态数据相结合,设计基于多编码器和捷联惯导的航位推算算法,该算法可以有效抑制航位推算误差的快速累积。在此基础上,采用扩展卡尔曼滤波算法对基于轮式里程计运动学模型解算的姿态数据和捷联惯导解算的姿态数据进行融合,通过提高航向精度来提高航位推算的定位精度。基于扩展卡尔曼滤波算法建立了捷联惯导/航位推算组合定位模型,针对建模不准确时扩展卡尔曼滤波存在滤波精度下降甚至发散的问题,提出了一种改进的自适应扩展卡尔曼滤波算法,通过实时估计并调整量测噪声来提高滤波的精度和稳定性。针对无轨胶轮车出现打滑或滑行等异常行驶状况导致的轮式里程计失效问题,建立了一种容错模型,实现了无轨胶轮车打滑或滑行等异常行驶状况的检测与隔离。搭建了移动小车实验系统,并将UWB定位结果作为参考真值,开展了综合跑车实验研究。实验结果表明:所提出的组合定位方法能够满足煤矿井下无轨胶轮车定位需求。Aiming at the problems of low accuracy and poor reliability of traditional trackless rubber tire vehicle positioning and orientation technology in coal mines,a trackless rubber tire vehicle positioning method based on the fusion of multiple encoders and inertial navigation is proposed.Based on the characteristics of multi encoder data,a kinematic model of a trackless rubber wheel odometer was established,and a heading calculation algorithm based on multi encoder was derived.In response to the problem of positioning divergence caused by relying solely on encoder data in the aforementioned dead reckoning algorithm,a dead reckoning algorithm based on multiple encoders and strapdown inertial navigation is designed by combining wheel odometry with attitude data.This algorithm can effectively suppress the rapid accumulation of dead reckoning errors.On this basis,the extended Kalman filtering algorithm is adopted to fuse the attitude data calculated based on the kinematic model of the wheel odometer and the attitude data calculated by the strapdown inertial navigation,and improve the positioning accuracy of the heading calculation by improving the heading accuracy.Based on the extended Kalman filtering algorithm,a combined positioning model for strapdown inertial navigation/dead reckoning was established.To address the problem of decreased or even divergent filtering accuracy when modeling is inaccurate,an improved adaptive extended Kalman filtering algorithm was proposed.By estimating and adjusting measurement noise in real time,the accuracy and stability of the filtering were improved.A fault-tolerant model was established to detect and isolate abnormal driving conditions such as slipping or sliding of trackless rubber wheeled vehicles,which resulted in the failure of wheel odometers.We built a mobile car experimental system and used UWB positioning results as reference truth values to conduct comprehensive sports car experimental research.The experimental results show that the proposed combination positioning method

关 键 词：无轨胶轮车 航位推算 捷联惯导系统 组合定位 自适应扩展卡尔曼滤波 

分 类 号：TD525[矿业工程—矿山机电]