支持多源感知数据集成的奶牛场物联网监控系统设计

A cow farm IOT monitoring system supporting the multisource sensor data integration

【目的】利用物联网技术实现养殖环境、日常活动、奶牛体况等信息的实时监测是当前奶牛养殖业的热点研究问题,由于当前物联网应用中传输协议标准不统一等导致的感知终端兼容性差、数据不易共享等问题阻碍了奶牛物联网深入应用的发展。【方法】文章从奶牛场物联网应用需求和数据特点出发,围绕异构多源感知数据接入、集成和操作等关键问题,开发基于REST物联网资源架构的奶牛场物联网监控系统。通过分析感知终端与互联网的通信特点,开发基于中间件的虚拟网关服务接口,提供不同协议终端的数据接入和解析服务;通过分析奶牛场各种资源本体关系,基于OWL和五元素表述法构建奶牛场资源本体描述模型,实现异构感知数据的语义集成和上下文关系描述;通过分析资源基础操作需求,基于Restful Web API开发资源统一操作接口,为上层应用奶牛场物联网监控系统开发提供快速开发基础。【结果】通过对奶牛场物联网监测系统的应用结果表明:该系统可实现上层物联网系统对多源异构感知终端的数据接入、数据集成和统一操作。【结论】该研究解决了奶牛场中异构感知终端的数据集成问题,不同种类感知数据的集成为奶牛生产进一步分析应用提供了良好的数据基础。

[Purposes]The variety of transportation and data protocols for different sensors in dairy farms leads to a poor compatibility usage among heterogeneous terminals and blocked the sharing of sensor data.In order to solve the problems of heterogonous data access integration and process,a resource architecture of IOT based on REST mode was promoted to develop the dairy farms IOT monitoring system.[Methods]By analyzing the communication characters between sensor terminals and internet,a virtual gateway service interface was developed based on middleware to provide the function of heterogonous data access and decoding.By researching the relations among various kinds of resources ontology in dairy farms,a dairy farms resources ontology model was built based on the OWL and the five elements description method to describe the semantic context and integrate the data.By analyzing the requirement for basic resources process,the Restful Web API was used to develop the unified process interfaces,which were the foundation of rapid development for upper IOT application system.The dairy farms IOT monitoring system was developed and test in a large-scale dairy farm in Hulunber,Northeast China.Sensors from three different companies were used in the system to capture the data of environment and cow health.[Results]The system has integrated the environment data and cow health data together and showed in the cow health record research system.[Conclusion]The integration data may provide a good data foundation for further application of intelligent cow farms management.