摘要
全球定位系统GPS提供了完善的室外定位服务,然而室内环境的多样性和复杂性,导致这一定位方式并不适用于室内。针对复杂的室内环境,提出一种基于深度学习的参考信号优化算法,提高室内蓝牙定位系统的精度。通过高斯滤波优化RSSI分布,利用深度学习方法优化参考信号,得到基准信号优化模型。实验结果表明,在10 m距离内,深度学习算法相对于传统平均方法具有更好的稳定性,并且能够有效地降低总体误差,为室内定位提供了重要的解决方案。
The Global Positioning System(GPS)provides a complete outdoor positioning service,but the diversity and complexity of the indoor environment make this positioning method unsuitable for indoor use.Aiming at the complex indoor environment,a reference signal optimization algorithm based on deep learning is proposed to improve the precision of indoor bluetooth positioning system.Gaussian filter is used to optimize RSSI distribution,and deep learning method is used to optimize reference signal,so as to get the reference signal optimization model.Results show that,in the range of 10 m,the deep learning algorithm has better stability than the traditional averaging method,and can effectively reduce the overall errors,which provides an important solution for indoor positioning.
作者
魏军
罗恒
倪启东
陈明哲
WEI Jun;LUO Heng;NI Qidong;CHEN Mingzhe(School of Electronic&Information Engineering,SUST,Suzhou 251009,China;Jiangsu Province Key Laboratory of Intelligent Building Energy Efficiency,Suzhou 215009,China)
出处
《苏州科技大学学报（自然科学版）》
CAS
2023年第2期78-84,共7页
Journal of Suzhou University of Science and Technology（Natural Science Edition）
基金
国家自然科学基金资助项目(61602334,51874205,51973109)。
关键词
室内定位
蓝牙定位
RSSI
深度学习算法
参考信号
indoor positioning
bluetooth positioning
RSSI
deep learning algorithms
reference signal
