Design And Implementation Of An Intelligent Decision-making Management System For Soilless Culture Based On The Internet Of Things

China is a big country with saline-alkali land,and the per capita arable land area is far below than the world average.Using soilless cultivation technology to improve land production efficiency has become an important way for Chinese agriculture to develop towards high efficiency.Promoting the application of Internet of Things technology in soilless cultivation and improving agricultural intelligent management are effective ways to accelerate agricultural modernization.The application of multi-sensor data fusion technology plays a key role in realizing low-cost and high-efficiency data collection and sorting.Due to the late starting of China’s agricultural Internet of Things,the technology and resources are not perfect enough,the accuracy of data fusion decision-making is low,and many systems lack of integration and information management.This paper designs and implements a comprehensive system of soilless cultivation data fusion decision model and information management cloud platform through field research,demand analysis,algorithm improvement and innovation.The specific research contents are as follows.(1)Analyzing the development background,application status and fusion decision-making methods of intelligent management in the environment of soilless cultivation and facility agriculture.On the basis of existing problems and actual needs,the system’s business processes and functional requirements are analyzed.Relying on the Io T,Web,and data fusion algorithms,the overall design of the system is carried out,and the overall system architecture is formulated.(2)Taking the soilless cultivation greenhouse environment as the background,a decision model based on multi-data fusion algorithm is constructed.The model is divided into two parts:first-level data fusion and global decision fusion.In the first-level data fusion stage,based on adaptive trust estimation,we propose a multi sensor data fusion algorithm in Trust Neural Network(T-NN),aiming to solve the problem of low accuracy and poor stability of multi-sensor data fusion.The time factor is introduced into the adaptive trust estimation model,by calculating the trust degree between nodes and optimizing the data,the data inaccuracy problem caused by the long measurement time is avoided to the greatest extent.The optimized data is introduced into the BP neural network for data fusion,which improves the accuracy and stability of the fusion.The optimized data is introduced into the BP neural network for data fusion,which improves the accuracy and stability of the fusion.In the global decision-making stage,the D-S evidence theory method to improve the evidence source is introduced to make decision fusion on the data after the first-level fusion optimization,which realizes the adaptive correction under the conflict of evidence and improves the accuracy of decision fusion.Through the comparison of simulation experiments,the accuracy of data fusion decision-making under this model is significantly improved,which provides a strong theoretical basis for the intelligent monitoring and decision-making control of the system.(3)Determining the selection of the system acquisition control terminal,and completing the design and implementation of the information management cloud platform.The selection of various sensors,controllers and other equipments required by the system has laid the foundation for the system construction.Through the analysis of the business process and functional requirements of the system,an intelligent decision management cloud platform based on Django+Vue is designed and developed.The platform designs a front desk for farm employees and a backstage management for administrators,and applies the data fusion decision model to the functional modules to realize intelligent monitoring of the shed environment,decision-making control,sensor anomaly detection,information and data management,and so on.After system deployment and testing,each module has reached the design requirements,which provides efficient operation management services for the greenhouse.