改进EKF-SLAM自动落布车定位算法  

An improved EKF-SLAM automatic dropping cloth vehicle positioning algorithm

作　　者：沈丹峰[1] 尚国飞 赵刚 柏顺伟 付茂文 SHEN Danfeng;SHANG Guofei;ZHAO Gang;BAI Shunwei;FU Maowen(School of Mechanical and Electrical Engineering,Xi’an Polytechnic University,Xi’an 710048,China;Shaanxi ChangLing Textile Mechtronical&Technology Co.Ltd.,Baoji 721013,Shaanxi,China)

机构地区：[1]西安工程大学机电工程学院,陕西西安710048 [2]陕西长岭纺织机电科技有限公司,陕西宝鸡721013

出　　处：《西安工程大学学报》2022年第6期77-85,共9页Journal of Xi’an Polytechnic University

基　　金：陕西省科技厅重点研发计划(S2022-YF-YBGY-0261)。

摘　　要：为提高自动落布车定位精度,解决扩展卡尔曼滤波(extended Kalman filter, EKF)同步建图与定位(simultaneous localization and mapping, SLAM)算法在强非线性系统中误差增大、系统定位精度降低等问题。结合多新息(muti-innovation)理论,提出迭代多新息扩展卡尔曼滤波同步建图与定位(IMI-EKF-SLAM)算法,使自动落布车定位系统在状态更新过程中,从对当前时刻的一次更新到对多个时刻的多次迭代更新,并对迭代值n、新息维度p进行了探讨。仿真和实验结果表明:改进算法减少了状态估计误差,并且当迭代值n为2,新息维度p为3时,IMI-EKF-SLAM算法定位效果最佳。In order to improve the positioning accuracy of the automatic dropping cloth vehicle, the problem of increasing error and decreasing positioning accuracy in the strongly nonlinear system is solved by the extended Kalman filter simultaneous localization and mapping(SLAM) algorithm. In this paper, an integrated muti-innovation extended Kalman filter simultaneous localization and mapping(IMI-EKF-SLAM) algorithm was proposed, which enabled the automatic dropping cloth vehicle localization system to change from a single update to multiple iterations in the process of state update, and explored the iteration value n and the new interest dimension p. The simulation and experimental results show that the improved algorithm reduces the state estimation error and the IMI-EKF-SLAM algorithm locates best when the iteration value n is 2 and the new interest dimension p is 3.

关 键 词：扩展卡尔曼滤波(EKF) 同步建图与定位(SLAM) 多新息理论 非线性系统 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置]