Design And Implementation Of Irrigation Control Platform Based On Microservice Architecture

As the development of agricultural IOT technology,the irrigation control equipment in agricultural production increases both in quantity and variety.Meanwhile,the management of the irrigation control equipment is also increasingly harder.Moreover,the coupling degree of traditional single architecture software platform is too high to satisfy the frequent iteration of modern agricultural system in function.Therefore,from the point of view of the practical demands,this paper studies the design and construction of the irrigation control platform by virtue of the micro-service architecture model and relevant technology modules.By analyzing the system requirements,this paper designs the fundamental component services with the Spring Cloud framework.For example,this paper completes the design of the registry service through Eureka and the design of routing gateway service through Zuul.Secondly,following the business functions of the irrigation control platform,this paper categorizes micro-services and completes the design of five business services,including the data display service and the control management service.In addition,this paper designs the business service architecture within the Spring Boot framework and through the MVC model.In this design,the quasi-realtime transmission of information is realized with Active MQ as an information intermediary.Next,this paper establishes a predictive model based on BP neural network algorithm to analyze the prediction of irrigation.In order to solve the problems of the initial weight of BP neural network and threshold randomness,this paper adopts the Artificial Bee Colony(ABC)to optimize BP neural network,on the basis of which the CNABC-BP algorithm comes into being.In this paper,the accuracy of the predictive model is verified by experimental simulation.In addition,this paper carries out the test of detailed functions of microservices components and business function modules,then analyzes test results.The test results of the irrigation control platform indicate that the micro-service architecture-based irrigation control platform in this paper has advantages including stable operation,user-friendly interface and smooth function development,and the error between the predicted irrigation amount and the actual result is only 7%,which permits the platform to meet the needs of projects and realize their goals.Thus,the platform is a valuable tool in multiple applications in reality.