Research On Big Data Management System For Yellow Tea Growth

China’s tea planting has a long history,and covering the whole country.However,tea planting methods have remained unchanged for many years,which makes it difficult to increase tea production.With the development of agricultural informationization,the real-time data acquisition capability has been greatly improved,which farmers can change the way of tea planting based on experience alone.In order to have a better understanding of tea planting,more and more data acquisition devices are deployed in tea plantations to obtain tea growth related data.How to acquire knowledge and rules from a large amount of data to improve tea planting methods has become an important research direction.In this paper,the environmental data and soil data of the growth process of yellow tea were obtained by deploying sensors,and the daily operation process of tea farmers was recorded.On this basis,a data management system was constructed.The system takes Jinyun Yellow Tea as the research object,and provides real-time data display,abnormal environment alarm and agricultural operation management functions for tea plantations.Furthermore,Optimizing the Method of Data Storage and Anomalous Data Processing in System.This paper mainly does the following work:(1)According to the requirement that expansibility and maintainability of yellow tea data management system,traditional monomer architecture can not meet the requirement,so micro-service is chosen as the architecture of the system.Analyze micro-service related technologies and research the process of microservice system implementation.(2)According to the problems that the single database increased storage pressure and slow data query which caused by rapid data increase,a hierarchical storage structure based on the data migration strategy of high-low water level algorithm with maximum storage is applied,which stores the data in different databases.After verification,the data migration effect is good,and the data query efficiency is greatly improved compared with the previous single database storage.(3)According to the problem that the noise data in the original data,an improvedDBSCAN algorithm is proposed for anomaly data processing.By calculating the Euclidean distance between all the data points and sorting the constructed distance matrix,obtaining the distance ascending curve of each row matrix,on the basis of this,the optimal range of values of the parameter Eps can be estimated.Through experimental verification,the parameter clustering effect within the estimated range is better.There is a good filtering effect,which proves that the improved algorithm is effective.(4)Design and implementation the functional modules of yellow tea big data management system.This paper analyzes the needs of tea gardens and designs the overall system.Research on how to implement the system function module and how to the construction of database.Through show the system function and Verify the effectiveness of the algorithm in the system,it proves the feasibility and effectiveness of the yellow tea growth data management system in this paper.