基于神经网络的磁场定位技术

doi:10.3969/j.issn.1000-1565.2022.06.014

摘要/Abstract

摘要： 由于磁场无处不在的特点,使得磁场定位广泛应用于目标定位和状态检测中,然而在含有复杂铁磁质环境下,磁场信号的变化会导致定位精度下降甚至不能定位等问题.针对上述问题,提出了一种将磁场定位与BP神经网络相结合的方法,并进行了实验验证. 结果表明,基于BP神经网络的磁场定位方法可用于含有铁磁质的复杂环境定位. 定位精度与数据采集时磁源的移动步长、磁场传感器数量及传感器电子噪声有关,移动步长越小,传感器数量越多,电子噪声越小,定位精度越高.

关键词: 磁场, 传感器, 定位, BP神经网络, 定位精度

Abstract: The ubiquitous nature of magnetic fields makes magnetic field localization widely used in target localization and condition detection. However, in the environment containing complex ferromagnetic mass, the variation of the magnetic field signal can lead to problems such as degradation of localization accuracy or even failure to localize. To address these problems, a method combining magnetic field localization with BP neural network is proposed, and experimental verified. The results show that the BP neural network-based magnetic field localization method can be used for the localization of complex environments containing ferromagnetic materials. The localization accuracy is related to the movement step of the magnetic source, the number of magnetic field sensors and the electronic noise of the sensors during data acquisition. The smaller the movement step, the more the sensors and the smaller the electronic- DOI:10.3969/j.issn.1000-1565.2022.06.014基于神经网络的磁场定位技术王华英^{1, 2, 3},孙海军¹,张雷^1,2,3,王学^1,2,3,黄艳宾^1,2,3,郭海军^1,2,3(1.河北工程大学数理科学与工程学院,河北邯郸 056038;2.河北省计算光学成像与光电检测技术创新中心,河北邯郸 056038;3.河北省计算光学成像与智能感测国际联合研究中心,河北邯郸 056038)摘要:由于磁场无处不在的特点,使得磁场定位广泛应用于目标定位和状态检测中,然而在含有复杂铁磁质环境下,磁场信号的变化会导致定位精度下降甚至不能定位等问题.针对上述问题,提出了一种将磁场定位与BP神经网络相结合的方法,并进行了实验验证. 结果表明,基于BP神经网络的磁场定位方法可用于含有铁磁质的复杂环境定位. 定位精度与数据采集时磁源的移动步长、磁场传感器数量及传感器电子噪声有关,移动步长越小,传感器数量越多,电子噪声越小,定位精度越高.关键词:磁场;传感器;定位;BP神经网络;定位精度中图分类号:TP394.1 文献标志码:A 文章编号:1000-1565(2022)06-0657-08Research on localization of magnetic field based on neural networkWANG Huaying^1,2,3, SUN Haijun¹, ZHANG Lei^1,2,3, WANG Xue^1,2,3, HUANG Yanbin^1,2,3, GUO Haijun^1,2,3(1. College of Mathematical Science and Engineering, Hebei University of Engineering, Handan 056038, China; 2. Hebei Computational Optical Imaging and Photoelectric Detection Technology Innovation Center, Handan 056038, China; 3. Hebei International Joint Research Center for Computational Optical Imaging and Intelligent Sensing, Handan 056038, China)Abstract: The ubiquitous nature of magnetic fields makes magnetic field localization widely used in target localization and condition detection. However, in the environment containing complex ferromagnetic mass, the variation of the magnetic field signal can lead to problems such as degradation of localization accuracy or even failure to localize. To address these problems, a method combining magnetic field localization with BP neural network is proposed, and experimental verified. The results show that the BP neural network-based magnetic field localization method can be used for the localization of complex environments containing ferromagnetic materials. The localization accuracy is related to the movement step of the magnetic source, the number of magnetic field sensors and the electronic noise of the sensors during data acquisition. The smaller the movement step, the more the sensors and the smaller the electronic- 收稿日期:2022-03-28 基金项目:国家自然科学基金资助项目(62175059);河北省创新能力提升计划资助项目(20540302D);邯郸市科学技术与发展计划项目(21422111246;19422031008-4) 第一作者:王华英(1963—),女,河北涉县人,河北工程大学教授,博士,主要从事机器视觉方面研究.E-mail: pbxsyingzi@126.com 通信作者:郭海军(1967—),男,河北广平人,河北工程大学实验师,主要从事机器视觉方面研究.E-mail: ghj6028039@163.com第6期王华英等:基于神经网络的磁场定位技术noise, the higher the localization accuracy.

Key words: magnetic field, sensor, positioning, BP neural network, positioning accuracy

中图分类号:

TP394.1

王华英,孙海军,张雷,王学,黄艳宾,郭海军. 基于神经网络的磁场定位技术[J]. 河北大学学报(自然科学版), 2022, 42(6): 657-664.

WANG Huaying, SUN Haijun, ZHANG Lei, WANG Xue, HUANG Yanbin, GUO Haijun. Research on localization of magnetic field based on neural network[J]. Journal of Hebei University(Natural Science Edition), 2022, 42(6): 657-664.

参考文献

[1] 张秀良,曲红.基于BP神经网络的计算机图像智能识别方法[J].信息与电脑(理论版), 2021, 33(15): 158-161.
[2] 吴鑫,赵红霞,罗筱毓,等.基于BP神经网络的岩石损伤声发射事件源定位研究[J].中国安全生产科学技术, 2021, 17(8): 36-42.
[3] 张靓.煤矿精确人员定位系统研究与设计[J].山西电子技术, 2021(4): 88-90.
[4] ZHAO L Z, WEN X B, DAN L. Amorphous localization algorithm based on BP artificial neural network[J]. Int Conf Softw Intell Technol Appl Int Conf Front Internet Things 2014, 2014: 178-183. DOI:10.1049/cp.2014.1556.
[5] 衣治安,马莉,刘延军.BP神经网络在无线传感网实时定位中的研究[J].电子技术与软件工程, 2014(17): 59.
[6] SUN D S, WEI E H, YANG L, et al. Improving fingerprint indoor localization using convolutional neural networks[J]. IEEE Access, 8: 193396-193411. DOI:10.1109/ACCESS.2020.3033312.
[7] 胡超,宋霜,阳万安,等.胶囊内窥镜位置方向磁场定位技术的研究[J].集成技术, 2012, 1(1): 105-113.
[8] 李修寒,竺明月,印佳,等.基于永磁体标记的胶囊内镜旋转测量定位模型设计[J].生物医学工程学进展, 2018, 39(2): 63-68. DOI:10.3969/j.issn.1674-1242.2018.02.001.
[9] 李泰宇,旷俭,牛小骥.基于磁强计阵列的室内行人定位算法研究[J].传感技术学报, 2020, 33(10): 1425-1431.
[10] 张宁,王三胜,易忠,等.基于全张量磁场梯度的磁偶极子定位及误差分析[J].航天器环境工程, 2017, 34(3): 317-323.DOI:10.3969/j.issn.1673-1379.2017.03.016.
[11] 陈雅娟,周子健,李清华,等.基于磁场矢量相位差的多传感器定位技术研究[J].航空科学技术, 2021, 32(10): 74-79. DOI:10.19452/j.issn.1007-5453.2021.10.012.
[12] 周家鹏,汪云甲,李昕,等.地磁室内定位技术研究[J].测绘通报, 2019(1): 18-22. DOI:10.13474/j.cnki.11-2246.2019.0004.
[13] KUSCHE R, SCHMIDT S O, HELLBRÜCK H. Indoor positioning via artificial magnetic fields[J]. IEEE Trans Instrum Meas, 2021, 70: 1-9. DOI:10.1109/TIM.2021.3073327.
[14] CHEN C H, CHEN P W, CHEN P J, et al. Indoor positioning using magnetic fingerprint map captured by magnetic sensor array[J]. Sensors(Basel), 2021, 21(17): 5707. DOI:10.3390/s21175707.
[15] BIAN S, HEVESI P, CHRISTENSEN L, et al. Induced magnetic field-based indoor positioning system for underwater environments[J]. Sensors(Basel), 2021, 21(6): 2218. DOI:10.3390/s21062218.
[16] HUANG X C, WANG W, LV Z P, et al. Precise positioning of the underground power cable by magnetic field detectio[C] //2019 IEEE Conference on Electrical Insulation and Dielectric Phenomena, Richland, WA, USA, IEEE, 702-705. DOI:10.1109/CEIDP47102.2019.9010585.
[17] SON W, CHOI L. Magnetic vector calibration for real-time indoor positioning[C] //ICC 2020 - 2020 IEEE International Conference on Communications. Dublin, Ireland, IEEE,: 1-7. DOI:10.1109/ICC40277.2020.9148955.
[18] 肖军.基于BP网络的船舶磁性目标定位方法研究[J].舰船科学技术, 2017, 39(24): 100-102.
[19] USTEBAY S, YINER Z, AYDIN M A, et al. An approach for evaluating performance of Magnetic-field based indoor positioning systems: neural network[C] //Comput Netw, 2017: 412-421. DOI:10.1007/978-3-319-59767-6_32.
[20] CHIANG T H, SUN Z H, SHIU H R, et al. Magnetic field-based localization in factories using neural network with robotic sampling[J]. IEEE Sens J, 2020, 20(21): 13110-13118. DOI:10.1109/JSEN.2020.3003404.
[21] ASHRAF I, KANG M Y, HUR S, et al. MINLOC: magnetic field patterns-based indoor localization using convolutional neural networks[J]. IEEE Access, 8: 66213-66227. DOI:10.1109/ACCESS.2020.2985384.
[22] 蒲春,孙政顺,赵世敏.Matlab神经网络工具箱BP算法比较[J].计算机仿真, 2006, 23(5): 142-144.
[23] 焦李成,杨淑媛,刘芳,等.神经网络七十年:回顾与展望[J].计算机学报, 2016, 39(8): 1697-1716.
[24] 蔡兵.BP神经网络隐层结构的设计方法[J].通化师范学院学报, 2007, 28(2): 18-19. (