基于窄带物联网通信技术的气象信息实时自动监测系统

A real-time automatic monitoring system for meteorological information based on narrowband Internet of Things (NB IoT) communication technology

摘要:

设计基于窄带物联网(NB-IoT)通信技术的气象信息实时自动监测系统,提升气象监测的实时性、准确性、覆盖范围,推动智慧气象的发展,提升灾害应对能力。构建基于NB-IoT的气象信息实时自动监测系统框架,数据感知层的数据采集节点利用不同类型传感器获取气象监测信息,由主控制器完成信息格式转换等处理后,依据预先定义的通信协议实现气象监测信息的打包,并上传给NB-IoT模块后,通过无线接入方式与数据传输层的NB-IoT网络建立连接,将气象监测信息传输给数据应用层,气象预测模块调用有序加权平均算子处理接收到的气象监测信息,将融合后的数据作为Storm流框架下的在线序列极限学习机模型的输入,输出气象预测结果,通过界面层呈现气象监测结果。实验结果表明：该系统的气象监测信息采集曲线与实际曲线贴合度高；气象预测的RMSE、MAE指标最低,分别为0.065、0.106；可实现异常气象监测信息预警；气象监测总功耗低、气象监测覆盖面广、稳定性高。

Abstract:

Design a real-time automatic monitoring system for meteorological information based on narrowband Internet of Things (NB IoT) communication technology, improve the real-time, accuracy, and coverage of meteorological monitoring, promote the development of smart meteorology, and enhance disaster response capabilities. Build a real-time automatic monitoring system framework for meteorological information based on NB IoT. The data collection nodes of the data perception layer use different types of sensors to obtain meteorological monitoring information. After the main controller completes information format conversion and other processing, meteorological monitoring information is packaged according to predefined communication protocols and uploaded to the NB IoT module. After that, the meteorological monitoring information is connected to the NB IoT network of the data transmission layer through wireless access, and transmitted to the data application layer. The meteorological prediction module calls the ordered weighted average operator to process the received meteorological monitoring information. The fused data is used as input for the online sequence extreme learning machine meteorological prediction model under the Storm flow framework, and the meteorological prediction results are output. The meteorological monitoring results are presented through the interface layer. The experimental results show that the meteorological monitoring information collection curve has a high degree of fit with the actual curve; The RMSE and MAE indicators for meteorological forecasting are the lowest, with values of 0.065 and 0.106, respectively; Can achieve abnormal meteorological monitoring information warning; Low total power consumption, wide coverage, and high stability in meteorological monitoring.