摘要
视障群体公交出行时,存在对精确引导登车的需求。基于公交车站特殊环境,常用接收信号强度指示(RSSI)测距技术进行测距定位。该测距技术具有成本低、复杂程度较小的特点,但在预测距离时易受环境干扰,导致测距误差偏大。提出采用粒子滤波(PF)模型对信号强度值进行预处理,再用反向传播(BP)神经网络进行距离精准预测,进而构建盲人公交登车引导系统。实验结果表明,提出的PF-BP神经网络模型测距精度比传统信号传播损耗模型平均提高66.3%,平均误差为0.373 m,且稳定在0.5 m以内,有效提升了测距定位精度。该系统能够为城市无障碍智能公交的定位系统升级更新奠定技术基础。
When the blind groups travel by bus,there is a demand for precise guidance and boarding.Based on the special environment of the bus station,Received Signal Strength Indicator(RSSI)ranging technology is commonly used for ranging positioning.This ranging technology has the characteristics of low cost and low complexity,but it is susceptible to environmental interference when predicting the distance,resulting in large ranging errors.A Particle Filter(PF)model is proposed to preprocess the signal intensity value,and then the Back Propagation(BP)neural network is used to accurately predict the distance,and then the blind bus boarding guidance system is constructed.The experimental results show that the range accuracy of the proposed PF-BP neural network model is 66.3%higher than that of the traditional signal propagation loss model,and the average error is 0.373 m,which is stable within 0.5 m,which effectively improves the accuracy of range finding and positioning.This system can provide strong support for the upgrade of urban barrier-free smart public transportation system.
作者
谢栋城
秦泽
周辰泠
姚振兴
XIE Dongcheng;QIN Ze;ZHOU Chenling;YAO Zhenxing(Chang′an University,Xi′an 710064,China)
出处
《导航定位学报》
CSCD
2022年第1期103-109,共7页
Journal of Navigation and Positioning
基金
国家自然科学基金青年基金项目(52002030)
教育部人文社会科学基金青年基金项目(20XJCZH011)。
关键词
接收信号强度指示
粒子滤波
反向传播神经网络
测距精度
公交导盲
received signal strength indicator
particle filter
back propagation neural network
ranging accuracy
blind guide for public transportation
