基于步态分类辅助的虚拟IMU的行人导航方法

doi:10.11896/jsjkx.211200148

计算机科学 ›› 2022, Vol. 49 ›› Issue (6A): 759-763.doi: 10.11896/jsjkx.211200148

基于步态分类辅助的虚拟IMU的行人导航方法

杨涵¹, 万游¹, 蔡洁萱¹, 方铭宇¹, 吴卓超¹, 金扬¹, 钱伟行²

1 南京师范大学电气与自动化工程学院南京 210046
2 南京师范大学江苏省大型科学仪器开放实验室南京 210023

出版日期:2022-06-10 发布日期:2022-06-08
通讯作者: 钱伟行(61192@njnu.edu.cn)
作者简介:(21190117@njnu.edu.cn)
基金资助:
南京师范大学江苏省大型科学仪器开放实验室开放课题

Pedestrian Navigation Method Based on Virtual Inertial Measurement Unit Assisted by GaitClassification

YANG Han¹, WAN You¹, CAI Jie-xuan¹, FANG Ming-yu¹, WU Zhuo-chao¹, JIN Yang¹, QIAN Wei-xing²

1 School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210046,China
2 Jiangsu Open Laboratory of Major Scientific Instrument and Equipment,Nanjing Normal University,Nanjing 210023,China

Online:2022-06-10 Published:2022-06-08
About author:YANG Han,born in 2001,undergra-duate.Her main research interests include navigation technology and deep learning.
QIAN Wei-xing,born in 1981,Ph.D,associate professor,master supervisor.His main research interests include integrated navigation technology,robot engineering and artificial intelligence.
Supported by:
Jiangsu Open Laboratory of Major Scientific Instrument and Equipment,Nanjing Normal University.

摘要/Abstract

摘要： 当人体剧烈运动或撞到障碍物使足部惯性测量单元(Inertial Measurement Unit,IMU)超程时,行人导航系统无法有效地实现定位导航。针对此问题,提出了一种行人导航方法,构建由步态分类辅助的虚拟惯性测量单元(Virtual Intertial Mea-surement Unit,VIMU)。此方法采用基于注意力的卷积神经网络(CNN)对常见的行人步态进行分类,以实际IMU实时采集的行人腿部和足部的惯性数据作为训练和测试样本。针对不同步态,分别建立了对应的ResNet-GRU混合神经网络模型,并根据这些模型构建足部VIMU,在足部实际IMU超程的情况下进行定位。实验表明,此方法能够有效提高基于零速度更新的行人导航系统在行人剧烈运动或与障碍物碰撞时的性能,从而增强系统在复杂和未知地形中的适应性。综合步态下的定位误差约为路径总长的1.43%,满足军用和民用的精度要求。

关键词: ResNet-GRU神经网络, 步态分类, 基于注意力的卷积神经网络, 行人导航, 虚拟IMU构建

Abstract: Due to the degraded performance of pedestrian navigation system when foot-mounted IMU is out of range during vigo-rous activities or collisions,a novel pedestrian navigation method is proposed based on construction of virtual inertial measurement unit(VIMU) assisted by gait classification.Attention-based convolutional neural network(CNN) is introduced to classify the common gaits of pedestrian.Then the inertial data from pedestrian's thigh and foot is collected synchronously via actual IMUs as training and testing samples.For different gaits,the corresponding ResNet-gated recurrent unit(Resnet-GRU) hybrid neural network models are built.According to these models,virtual foot-mounted IMU is constructed for positioning in case of actual foot-mounted IMU overrange.Experiments show that,the proposed method brings enhanced performance of pedestrian navigation system based on zero velocity update when the foot motion of pedestrian is violent,which makes the navigation system more adaptable in complex and unknown terrains.The positioning error during comprehensive gait is about 1.43% of the total walking distance,which satisfies the accuracy requirement of military and civilian applications.

Key words: Attention-based convolu-tional neural network, Gait classification, Pedestrian navigation, ResNet-gated recurrent unit neural network, Virtual inertial measurement unit construction

中图分类号:

O643

杨涵, 万游, 蔡洁萱, 方铭宇, 吴卓超, 金扬, 钱伟行. 基于步态分类辅助的虚拟IMU的行人导航方法[J]. 计算机科学, 2022, 49(6A): 759-763. https://doi.org/10.11896/jsjkx.211200148

YANG Han, WAN You, CAI Jie-xuan, FANG Ming-yu, WU Zhuo-chao, JIN Yang, QIAN Wei-xing. Pedestrian Navigation Method Based on Virtual Inertial Measurement Unit Assisted by GaitClassification[J]. Computer Science, 2022, 49(6A): 759-763. https://doi.org/10.11896/jsjkx.211200148

参考文献

[1] HSU Y L,WANG J S,CHANG C W.A wearable inertial pedestrian navigation system with quaternion-based extended Kalman filter for pedestrian localization[J].IEEE Sensors Journal,2017,17(10):3193-3206.
[2] ABDELMOUMEN N,ZAKARIA K,JOERG B.Step detection for ZUPT-aided inertial pedestrian navigation system using foot-mounted permanent magnet[J].IEEE Sensors Journal,2016,16(17):6766-6773.
[3] WANG G C,XU X S,YAO Y Q,et al.A novel BPNN-based method to overcome the GPS outages for INS/GPS system[J].IEEE Access,2019,7:82134-82143.
[4] YAO Y Q,XU X S,ZHU C C,et al.A hybrid fusion algorithm for GPS/INS integration during GPS outages[J].Measurement,2017,103:42-51.
[5] HAMILL J,KNUTZEN K M.Biomechanical basis of humanmovement [M].Philadelphia:Lippincott Williams & Wilkins,2006.
[6] WANG L,ZANG J L,ZHANG Q L,et al.Action recognition byan attention-aware temporal weighted convolutional neural network[J].Sensors,2018,18:1979-1997.
[7] HU J,SHEN L,SUN G.Squeeze-and-Excitation Networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.2018:7132-7141.
[8] XIAO X,LI K.Multi-Label classification for power quality disturbances by integrated deep learning[J].IEEE Access,2021,9:152250-152260.
[9] FAN Q G,SUN Y,SUN B W,et al.Pedestrian indoor positioning system based on GLRT zero speed detection[J].Chinese Journal of Sensors and Actuators,2017,30(11):1706-1711.
[10] LIU J Y.Theory and application of navigation system [M].Xi'an:Northwestern Polytechnical University Press,2010.
[11] DENG Z,WANG P,YAN D,et al.Foot-Mounted pedestriannavigation method based on gait classification for three-dimensional positioning[J].IEEE Sensors Journal,2020,20(4):2045-2055.
[12] YANG S Q,XING L,LIU W H,et al.Robust navigation me-thod for wearable human-machine interaction system based on deep learning[J].IEEE Sensors Journal,2020,20(24):14950-14957.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于步态分类辅助的虚拟IMU的行人导航方法

Pedestrian Navigation Method Based on Virtual Inertial Measurement Unit Assisted by GaitClassification

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 0

[1]	周婧,陈庙红,吴豪杰. 基于惯性导航的平面航迹推算的研究 Research on Plane Dead Reckoning Based on Inertial Navigation System 计算机科学, 2017, 44(Z6): 582-586. https://doi.org/10.11896/j.issn.1002-137X.2017.6A.131
[2]	郑炜，王卫星，梁顺龙. 面向行人导航的RMLT DR算法研究与实现 RMLT DR Algorithm's Research and Implementation in Pedestrian Navigation System 计算机科学, 2011, 38(5): 240-243.