| 267 | 11 | 168 |
| 下载次数 | 被引频次 | 阅读次数 |
针对室内环境多变引起的地磁数据浮动、指纹库分辨率低导致误匹配概率大的问题,提出基于多种群遗传算法(Multi-Population Genetic Algorithm, MPGA)优化支持向量机(Support Vector Machine, SVM)的地磁室内定位算法。数据预处理阶段采用高斯滤波去噪,利用SVM构建地磁定位模型,选择高斯径向基核函数作为SVM的核函数,并通过MPGA的寻优性确定SVM最优参数。试验结果表明,该算法能较好地弥补现有算法存在的缺陷,有效提升定位精度及稳定性。
Abstract：To solve the problems of floating geomagnetic data and high mismatching probability caused by the variable indoor environment and the low resolution of fingerprint database, a multi-population genetic algorithm(MPGA) is proposed to optimize the support vector machine(SVM) for indoor geomagnetic positioning algorithm.In the data preprocessing stage, Gaussian filtering is used to denoise the data, and SVM is used to build the geomagnetic positioning model.The Gaussian radial basis kernel function is selected as the kernel function of SVM,and the optimal parameters of SVM are determined by the optimization of multi-population genetic algorithm.The experimental results show that the proposed algorithm can make up for the defects of the existing algorithms, and effectively improve the positioning accuracy and stability.
[1] 高伟，侯聪毅，许万旸，等.室内导航定位技术研究进展与展望[J].导航定位学报，2019,7(1):10-17.
[2] ASHRAF I,DIN S,HUR S J,et al.Empirical Overview of Benchmark Datasets for Geomagnetic Field-Based Indoor Positioning[J].Sensors (Basel),2021,21(10):1-10.
[3] DUONG B N,HE J,TRUNG T,et al.An Effective Random Statistical Method for Indoor Positioning System Using WiFi Fingerprinting[J].Future Generation Computer Systems,2020,109:238-248.
[4] ZHOU N,LAU L,BAI R,et al.Novel Prior Position Determination Approaches in Particle Filter for Ultra Wideband (UWB)-based Indoor Positioning[J].Navigation,2021,68(2):277-292.
[5] 陈珊珊，史志才，吴飞，等.粒子群优化人工神经网络的RFID室内定位算法[J].传感器与微系统，2018,37(12):136-138.
[6] 左肖，邵建华，沈宏杰，等.自然环境光下的室内可见光定位系统[J].激光杂志，2019,40(11):7-11.
[7] 屈阳，万国龙，丛丽，等.采用指纹SD权重的FM两级室内定位方法[J].导航定位学报，2018,6(3):36-40.
[8] SUN M,WANG Y J,XU S L,et al.Indoor Positioning Integrating PDR/Geomagnetic Positioning Based on the Genetic-Particle Filter[J].Applied Sciences,2020,10(2):1-5.
[9] 白艳茹，罗海勇，曹程琳，等.基于在线学习磁校准的行人航位推算算法[J].北京邮电大学学报，2021,44(3):53-60.
[10] 孙猛，汪云甲，周家鹏，等.一种基于快速动态时间规整的地磁定位算法[J].测绘科学，2020,45(8):77-82.
[11] KIM S H,KANG D H,KIN K W,et al.Indoor Location Estimation Using WiFi RSSI Signals and Geomagnetic Sensors[J].IEMEK Journal of Embedded Systems and Applications,2017,12(1):19-25.
[12] HE H,WEI L,AN L D,et al.An Improved Particle Filter Algorithm for Geomagnetic Indoor Positioning[J].Journal of Sensors,2018(24):1-9.
[13] MAHDI A,GRéGOIRE L.Improving Indoor Geomagnetic Field Fingerprinting Using Recurrence Plot-based Convolutional Neural Networks[J].Journal of Location Based Services,2021,15(1):61-87.
[14] 华胜.基于AHP-WKNN和改进AKF的室内定位算法[D].南京：南京邮电大学，2020.
[15] ZHANG T,MAN Y.The Enhancement of WiFi Fingerprint Positioning Using Convolutional Neural Network[C]//Proceeding of 2018 International Conference on Computer,Communication and Network Technology (CCNT 2018).Tongxiang:DEStech Publications,2018:226-230.
[16] ROY P,CHOWDHURY C.A Survey of Machine Learning Techniques for Indoor Localization and Navigation Systems[J].Journal of Intelligent & Robotic Systems,2021,101(3):1170-1182.
[17] 何清，李宁，罗文娟，等.大数据下的机器学习算法综述[J].模式识别与人工智能，2014,27(4):327-336.
[18] 宋宁佳，崔英花.基于GA-GRNN的RFID室内定位算法[J].计算机工程，2019,45(11):298-302.
[19] 李维，黄鹤，罗德安.改进粒子滤波器的室内地磁匹配精度研究[J].测绘科学，2018,43(7):109-114.
[20] 周家鹏.基于粒子滤波的地磁/PDR室内融合定位技术研究[D].徐州：中国矿业大学，2019.
基本信息:
中图分类号:TN92;TP18
引用信息:
[1]韩雨辰,吕伟才,仲臣,等.基于MPGA-SVM的地磁室内定位算法研究[J].无线电工程,2021,51(11):1313-1319.
基金信息:
安徽省自然科学基金资助项目(2008085MD114); 国家自然基金面上项目(41474026); 安徽省重点研究与开发计划(202104a07020014)~~
2021-07-05
2021
2021-08-19
2021-08-27
2021
1
2021-11-05
2021-11-05