传感技术学报2026，Vol.39Issue(3)：509-517,9.DOI:10.3969/j.issn.1004-1699.2026.03.007
基于分层自适应随机森林的无源RFID相对定位方案
Passive RFID Relative Localization Scheme Based on Hierarchical Self-Adaptive Random Forest
摘要
Abstract
For high-precision and low-cost relative localization of passive RFID tags in three-dimensional space,a relative positioning scheme based on the packet reception rate(PRR)measurement and the hierarchical self-adaptive random forest algorithm is proposed.As the beginning step,the PRRs of passive tags at observation points around the target area are collected to build the offline fingerprint database.Afterwards,two lightweight online models are obtained for matching the PRR measurement and the relative position,based on the hierarchical decision tree model and the further extended hierarchical self-adaptive random forest model,respectively.Meanwhile the hyper-parameters of the proposed models are optimized by the random grid search strategy.Practical experiments carried out in libraries and indoor corridors show that the accuracy can reach 98.04%in static environments and 91.1%in dynamic environments,outperfor-ming the existing other machine learning algorithms.关键词
无线传感器网络/室内定位/指纹匹配/无源RFID/相对位置/随机森林Key words
wireless sensor network/indoor location/fingerprint matching/passive RFID/relative location/random forest引用本文复制引用
武梅,靳乾坤,周彪..基于分层自适应随机森林的无源RFID相对定位方案[J].传感技术学报,2026,39(3):509-517,9.基金项目
国家自然科学基金项目(61701385) （61701385）
