Research And Realization Of Integrated Cloud System Based On Deep Learning And Internet Of Things

With the rapid development of machine learning and internet of things technology,as well as the development of urbanization in China,China's agriculture is welcoming new opportunities for development.Using artificial intelligence technology to effectively mine agricultural big data and achieve effective control and management of smart agriculture has become a research hotspot and a difficult point.In this paper,based on RNN recurrent neural network and internet of things technology,the intelligent agriculture water and fertilizer integration cloud system is proposed.Firstly,based on RNN circulant neural network,an integrated water and fertilizer control method was proposed to realize the dynamic control and precise control of water and fertilizer irrigation.The effectiveness of the integrated control method was validated by the indoor greenhouse model.The experimental results show that the RNN method is relative to artificial irrigation increased water and fertilizer utilization by about 30% and 17% respectively,and increased water and fertilizer by about 20% and30%.Secondly,an intelligent agricultural water and fertilizer integrated cloud system architecture focusing on the management of agricultural big data and integrated control of water and fertilizer was put forward.The basic structure of the system was defined through the analysis of operational requirements,functional requirements,and non-functional requirements of the cloud system.The overall design point of view gives the system architecture and system function modules,and each module is designed in detail and the database design is completed.Then,the water and fertilizer integration cloud system was implemented,which focused on the farmland moisture monitoring module,data aggregation and storage module,water and fertilizer integrated control module and agricultural data analysis module.Finally,this paper tests the system from three aspects: test case,system test and non-functional test.The test results are consistent with the expected results.