分布式多源农林物联网感知数据共享平台研发

由于农业和林业传感器种类繁多、数据传输协议多样,在各个物联网数据应用系统之间形成了信息孤单现象,难以实现物联网传感器数据的交互共享。因此,该文以农林领域常见传感器数据为研究对象,针对不同类型传感器节点及采集数据差异性等特点,采用符号表示法表达传感器节点数据,设计了通用数据交互格式;针对其海量性特点,采用分布式面向服务的结构方法及成熟的开发技术,设计了分布式多源农林物联网感知数据共享平台。该平台由数据中心子系统、数据适配子系统、数据存储子系统、数据发布子系统及数据传输总线5部分组成,分别实现了传感器节点注册、多源差异数据的接入适配、数据的分布式存储、数据标准化发布及数据传输等功能,为农林领域不同物联网设备与数据应用系统之间架起了数据桥梁,实现了农业和林业物联网感知数据的统一管理。目前,平台分别接入了顺义、新疆、杨凌、通州等17个合作单位的550个传感器节点数据,接入数据量每天超过10 000条,运行状态良好。

Development of distributed data sharing platform for multi-source IOT sensor data of agriculture and forestry

Abstract: Agriculture and forestry IOT sensor data are the basis of agricultural informatization and forest ecological monitoring. At present, the phenomenon of information alone between various IOT data application systems makes it difficult to realize IOT sensor data sharing. However, with the developing of agricultural whole industry chain information management and forest ecological monitoring domain, the demand for each link of IOT sensor data of agriculture and forestry are getting stronger. In order to overcome the problem, a sharing platform for distributed multi-source IOT sensor data of agriculture and forestry was designed. The platform was divided into 5 parts, including data center subsystem, data adapter subsystem, data storage subsystem, data publishing subsystem and data transmission bus. The data center subsystem was responsible for managing the basic information of the sensor, and realized basic information storage function and registration function of the sensor nodes. This subsystem was the core of the platform for data management, which provided basic information of the sensor nodes for other subsystems. The data adapter subsystem was responsible for receiving and analyzing data from different sensor nodes, and it used a non-blocking Socket interface and a data acquisition interface based on Http and WebService protocol to receive data from different data sources. In this subsystem, the adapter was designed according to the difference format of the received data, and the final data to be analyzed were sent to the data storage subsystem. The data storage subsystem was responsible for storing sensor nodes data using MySQL database. The data publishing subsystem provided standardized data query interfaces of sensor nodes, which included a single sensor node real-time data and historical data query interface, multi sensor nodes real-time data and historical data query interface. The data transmission bus was realized by Active MQ server software, and concurrent capability and response speed of the platform were tested, Concurrent capability test results showed that, error rate was 0 and data throughput was 2 724 KB when the data concurrent access link number was below level 102, which could meet the requirements of concurrent. Response speed test results of the platform showed that, response time of single sensor real-time data query was 63 ms, response time of less than 100 sensors real-time data query was 90.40 ms, average response time of historical data query was 1 846.72 ms when data were less than 1 000, and average response time of historical data query was 3 353.86 ms when data were more than 10 000, which met design requirements. The sharing platform has been deployed in laboratory of National Engineering Research Center for Information Technology in Agriculture, and the performance has been well since began to run. Currently, 550 sensor nodes'' data of Shunyi, Xinjiang, Yangling, Tongzhou and other 17 units have been accessed, with the amount of accessing data more than 10 000 every day. The platform sets up a data bridge between different IOT sensor devices and the data application system in agriculture and forestry field, and realizes the unification management of the agricultural and forestry IOT sensor data.