Creation of an Early Warning Platform for Risks to the Quality and Safety of Stored Grain

Abstract

Based on the fast detection of mycotoxins and heavy metals in grain by each storage enterprise of Beijing reserve grain, the real-time migration of monitoring data is realized by adopting the B/S architecture, terminal WEB browser, and using the wireless transmission function of the rapid detection equipment to collect and transmit the data through the encryption algorithm; by using P2MP point-to-multipoint network communication, the data collected by multiple devices online/offline are uploaded through the file uploading system and the relational database extraction tool to achieve synchronous integration of data from multiple sources; by using relational database MySQL, column storage database HBase and distributed file system HDFS, distributed full-text search ElasticSearch and distributed memory database Redis to achieve cloud storage; by using Storm for real-time computation, Streaming for streaming operation, Spark for memory computing, and MapReduce for batch computing to achieve rapid processing of monitoring data; by using artificial intelligence technology and MLlib/Mahout for data mining and modelling to form a spatial and temporal sequence model of Beijing's grain producing areas and credit evaluation model of grain purchasing and marketing enterprises, so as to achieve the dynamic early warning of grain quality and safety and visual expression of the data, and provide early warning judgments to facilitate real-time control by government departments and real-time response to the situation. It promotes real-time control, real-time response and traceability management of grain quality and safety by management departments, enhances healthy competition and the establishment of an integrity system in grain circulation and storage, and provides a scientific basis for government decision-making.

Keywords

• risk early warning
• grain
• fast detection
• information technology
• quality safety