信息物理系统事件驱动下的农业气象监测系统

Agro-meteorological monitoring system based on event-driven modeling of cyber-physical system

针对传统农业气象观测和当前传感器技术存在的不足,设计了一套基于信息物理系统(Cyber-physical system,CPS)的气象远程监测系统。针对信息物理系统时空特性,基于节点事件驱动方法,建立了3层信息物理系统模型,并给出了时空事件建模方法。该系统由边缘设备负责感知,云平台负责计算,通过事件-行为模式保证时空同步性,实现了气象信息采集、传输和处理的高度集成,同时在网络边缘处应用卷积神经网络实现设备电量识别以自适应采集频率。通过在河北工业大学测试点开展的采集试验和系统模型试运行结果表明,系统表现出较好的实时性、稳定性和时空同步性,农业气象信息的采集、传输、处理和远程监控等各项功能均可满足各级用户需求。

In view of the shortcomings of traditional agricultural meteorological observation and current sensor technology,a set of meteorological remote monitoring system based on cyber-physical system（ CPS） was designed. In view of the temporal and spatial characteristics of CPS,a three layer CPS model was established based on the node event-driven method and spatiotemporal event modeling method was put out. In this system,edge device was responsible for perception and cloud platform was responsible for the calculation. The system ensured concurrency in time and space by event-behavior model and completed the integration of acquisition,transmission and processing of meteorological information. Moreover,at the edge of the network,convolutional neural network was used to identify the system power and adaptively change the sampling frequency to minimize energy consumption of the system. The experiment and system model test run results at the test site of Hebei university of technology showed that the system had good real-time, stability and spatiotemporal synchronization. Besides, the functions of acquisition,transmission,processing and remote monitoring of agricultural meteorology can meet the needs of users at all levels.