The Research And Implementation Of Early Warning System Based On Agricultural Product Tracing Platform

Farming is a special,complex and independent work.It is also the most important work.It is related to human health and the whole ecological cycle.With the rapid development of agriculture in China,at the same time,the supervision of agricultural products has not reached a certain extent,the quality and safety of agricultural products occur frequently,which not only poses a great threat to people's daily life,but also seriously affects the growth of the whole society and economy.More and more consumers at different levels are also increasing demand for safety,health and quality assurance of agricultural products.The development of social informatization has brought effective supervision plan to supervise the production and sale process of agriculture.Using information technology,it is very important to construct a complete information chain,such as the sources,components,processing and flow information of agricultural products and their inputs,and the historical information of these products is called tracing information.In the existing tracing system,we use the threshold design method to detect abnormal data in tracing information.But for complex product direction,the threshold needed to design is more complex,and experts need to judge the threshold value.Based on this,this paper studies the early warning problem of abnormal product data in the agricultural product tracing system.On the basis of the traditional early-warning system,such as fruit and vegetable,meat and dairy,and so on,it combines Markov model and convolution neural network algorithm to design abnormal detection model,and develops the autonomous detection system of abnormal data.To achieve abnormal self detection and tracking in Agriculture product traceability platform.In view of different conditions,a threshold management system is created under the condition of considering the characteristics of different agricultural product data.A Markov model is established for a single data of a certain agricultural product to predict the next time point,and the abnormal early warning is carried out according to the probability judgment of different predicted values;according to the different products.The characteristic matrix is created between the flow point and the different products,and the data is judged according to the trend of the data in the matrix.The system uses the Spark+Kafka large data processing architecture,and the web development technology based on the Java language,and uses Mysql as the backstage database to realize the function of the module in the early warning system.With this framework,the real-time,high concurrency and high fault tolerance requirements of agricultural products early warning system are realized.This system has realized the abnormal early warning of agricultural products in the whole supply chain,aiming at reducing the unnecessary wastage of agricultural products,improving the efficiency of agricultural products anomaly detection and ensuring the quality and safety of agricultural products.The system provides advance warning and problem tracking for regulatory authorities with a perfect function of process control,information inquiry and management,and alarm prediction.