Recognizing the position of moving objects, facilities or important equipments has been regarded as a hot issue in the future environment which needs ubiquitous characteristics. Furthermore, it is indispensably needed...Recognizing the position of moving objects, facilities or important equipments has been regarded as a hot issue in the future environment which needs ubiquitous characteristics. Furthermore, it is indispensably needed to acquire location data in order to realize an infrastructure of sensor network system. Namely, distance data are the preliminary information to complete localization of a moving object in local area. Even though there have been several researches on this topic, there are however still much amount of problems in getting precisely exact distance data. Out of them, the research on using TOF (Time-of-flight) is very few as well as could be thought of as a cutting-edge technique. Moreover, it is very difficult to get minutely RF signal's TOF for getting short distance data; that is, very complicated signal processing skills are requested for realizing it. In this paper, we present a new methodology of RF short ranging technique based on vernier effect through using of heterogeneous operating clocks.展开更多
文摘Recognizing the position of moving objects, facilities or important equipments has been regarded as a hot issue in the future environment which needs ubiquitous characteristics. Furthermore, it is indispensably needed to acquire location data in order to realize an infrastructure of sensor network system. Namely, distance data are the preliminary information to complete localization of a moving object in local area. Even though there have been several researches on this topic, there are however still much amount of problems in getting precisely exact distance data. Out of them, the research on using TOF (Time-of-flight) is very few as well as could be thought of as a cutting-edge technique. Moreover, it is very difficult to get minutely RF signal's TOF for getting short distance data; that is, very complicated signal processing skills are requested for realizing it. In this paper, we present a new methodology of RF short ranging technique based on vernier effect through using of heterogeneous operating clocks.
文摘为提升法布里-珀罗(Fabry-Pérot,F-P)传感器游标光谱信号解调的准确性,提出基于深度学习的光谱数据直接解调方法。首先对光谱数据进行预处理,将复杂的游标光谱信息转化为卷积神经网络(Convolutional Neural Network,CNN)可以处理的数据格式,然后采用深度学习模型对预处理后的完整光谱数据进行训练和测试,并利用卷积神经网络对光谱数据进行特征提取和分类,最终实现待测信号的准确解调。使用灵敏度为112.5 nm/MPa的双腔法布里-珀罗传感器采集光谱数据,并开展信号解调实验,结果表明:CNN模型对未知光谱进行10折(fold)交叉验证的平均准确率为92.49%,均方根误差RRMSE(Root Mean Square Error,RMSE)为0.0392 MPa,相对误差的平均值为3.31%;卷积神经网络-长短期记忆(Convolutional Neural Network-Long Short Term Memory,CNN-LSTM)模型对未知光谱进行10折交叉验证的平均准确率为96.98%,RRMSE为0.0390 MPa,相对误差的平均值为3.28%。基于CNN-LSTM模型的方法仅通过解调256个采样点的数据就实现了较高准确度,具有便捷、高效的优点,为推动光谱信号解调领域发展提供了有效的技术途径,为开发智能光学传感系统提供了重要参考。
