Research And Design Of Intelligent Greenhouse Internet Of Things Platform Based On Raspberry Pi

Some regions and enterprises in China have adopted smart agriculture projects.But most of the front-end data acquisition nodes of the project use the single-chip architecture,lack of scalability,unit cost is too high,and the latest technical support for the frontier artificial intelligence is insufficient.In addition,high concurrency and high throughput are not fully considered for the server.This paper discusses in detail that the intelligent greenhouse Internet of Things platform should have functional characteristics.Based on the latest high-performance computing terminal of raspberry pie,a set of low-cost,superior performance and good scalability front-end data acquisition nodes is designed.Based on the popular Spring framework in Java development system,a high concurrency and high availability back-end service support platform is designed.The whole system is divided into raspberry data acquisition front-end module,Netty-based Internet of Things data transmission gateway,Spring-based Internet of Things big data service background,and Vue-based front-end integrated control platform.In this paper,the hardware design of data acquisition front-end machine based on raspberry pie is discussed in detail,and four sub-modules of data acquisition,data transmission and intelligent control based on decision-making system are analyzed and designed one by one.In order to improve the system’s data throughput and high concurrent accesses,a data transmission gateway module of Internet of Things based on Netty is designed.In order to make the back-end management module of the platform have the ability to extend to PC,mobile phone,tablet or other mobile terminals,this paper designs a front-end and back-end separated software architecture.The service side uses Spring framework + RESTful style interface mode,and the front-end uses SPA application based on Vue.js framework,and can be adapted to various mobile terminals.A complete test scheme is designed for each sub-module of the whole system,and various automatic test tools are applied to fully verify the high reliability and concurrency of the system.Finally,the core decision-making system of intelligent greenhouse is simulated and tested,and the two algorithms are compared.It is concluded that the way of fuzzy decision-making system is more rapid.The decision-making system and the manual control method are tested,which proves that the intelligent platform can control the temperature and humidity stably in the greenhouse.