摘要
针对地下停车场、地下矿井等用户场景定位服务的高精度需求,设计了一种基于WiFi信号的室内定位系统.系统的设计引入了深度神经网络算法,对WiFi指纹数据进行训练,得到一种室内定位模型.通过对UJIIndoorLoc数据进行实验仿真,结果表明,该室内定位模型的楼层定位准确率较传统机器学习模型提升约6%,位置定位精度较传统模型提升约7%.通过对实际应用场景进行测试,测试结果表明该室内定位模型的定位精度优于传统机器学习模型,提升约5%.
For the high requirements of location service such as underground parking lots and underground mines,an indoor positioning system based on WiFi signals is designed.The indoor positioning system introduces a deep neural network algorithm to train the WiFi fingerprint data and to obtain an indoor positioning model.Through the experimental simulation of UJIIndoorLoc data,the simulation results show that floor positioning accuracy of the indoor positioning model is about6% higher than that of the traditional machine learning model,and the positional positioning accuracy is about 7% higher than that of the traditional model.Through the test of the actual application scenario,the test results show that the positioning accuracy of the indoor positioning model is better than that of the traditional machine learning model,which is about 5%.
作者
梁冀
吴彬
LIANG Ji;WU Bin(College of Physics and Electronic Engineering ,Guangxi Normal University for Nationalities ,Chongzuo 532200,China;China Mobile Communications Group Guangxi Co.Ltd.,Nanning 530000,China)
出处
《内蒙古大学学报（自然科学版）》
CAS
北大核心
2019年第2期199-204,共6页
Journal of Inner Mongolia University：Natural Science Edition
基金
国家自然科学基金项目(61562006)
2015年度广西高校科研项目(KY2015LX542)
广西民族师范学院科研经费资助项目(2018YB027)
