摘要
目前室内定位技术仍处于探索阶段,正在研究的室内定位技术包括WiFi、Bluetooth、RFID、Zigbee、可见光定位、地磁定位、红外线定位和超声波定位等,通过这些独立的技术手段已经实现5~10 m的室内定位精度,但是在工作过程中这些技术手段往往存在可靠性、连续性和稳定性不足的现象,难以满足大众位置服务和应急救援的需求。在研究智能手机的WiFi、蓝牙和地磁等多传感器的基础上,提出了一种混合室内云定位技术,通过基于深度学习的多层指纹特征提取技术、基于众包数据的指纹数据更新技术,最终实现可靠连续稳定的定位结果,达到优于3 m的室内定位精度。
At present,indoor positioning technology is still in the exploratory stage.The indoor positioning technologies being studied include WiFi,Bluetooth,RFID,Zigbee,visible light positioning,geomagnetic positioning,infrared positioning,and ultrasonic positioning,etc.Through these individual technical means,an indoor positioning accuracy of 5~10 m has been achieved.While in the operation process of these technologies,problems such as lack of reliability,continuity and stability often exist,which makes it difficult to meet the needs of public location services and emergency rescue.Based on the research of multiple sensors of smartphones such as WiFi,Bluetooth,and geomagnetism,a hybrid indoor cloud positioning technology is proposed.Through the multi-layer fingerprint feature extraction technology based on deep learning and the fingerprint data update technology based on crowdsourcing data,reliable,continuous and stable indoor positioning results are obtained,and an indoor positioning accuracy better than 3 m is achieved.
作者
梁晓虎
甘兴利
张衡
黄璐
LIANG Xiaohu;GAN Xingli;ZHANG Heng;HUANG Lu(The 54th Research Institute of CETC,Shijiazhuang 050081,China;State Key Laboratory of Satellite Navigation System and Equipment Technology,Shijiazhuang 050081,China)
出处
《无线电工程》
2020年第2期108-112,共5页
Radio Engineering
基金
国家“十三五”重点研发计划基金资助项目(2016YFB0502100)”~~
关键词
多传感器
室内定位
深度学习
指纹数据
云定位
multi-sensor
indoor positioning
deep learning
fingerprint data
cloud positioning
