Research And Application On Agricultural Environmental Data Collection Based On Deep Learning And Beidou Short Message

Our country is a large agricultural country,and agriculture is the foundation of the national economy.The sustainable and comprehensive development of agriculture is related to the coordination of national food security and the national economy.The agricultural Internet of Things,as a combination of Internet of Things technology and agriculture,promotes the development of digitalization and intelligence in agriculture,among which agricultural environmental data is the cornerstone of the agricultural Internet of Things.Driven by the "village revitalization" strategy,the agricultural Internet of Things will be promoted in rural areas to promote agricultural transformation and upgrading and sustainable development in rural areas.In remote rural areas,facing environmental factors such as weather and communication problems,agricultural Io T devices may not be able to collect agricultural environmental data stably and reliably.In order to solve this problem,based on deep learning and Beidou short message technology to study agricultural environmental data collection,the research content and results of this paper are as follows:(1)Aiming at the problem that agricultural environmental data collection equipment cannot continue to collect steadily due to the weather,an intelligent collection mode based on deep learning is proposed.Firstly,an energy consumption model is established for the three parts of the sensor,controller and communication module of the environmental data collection node;then,the SSAE-Bi LSTM photovoltaic power generation power prediction model is proposed.The model uses a variety of factors affecting photovoltaic power generation power as input characteristics,through SSAE The network automatically extracts low-dimensional abstract features from the input,and uses the abstract features as the input of the Bi LSTM network to achieve short-term prediction of photovoltaic power generation;finally,based on the energy consumption model and photovoltaic power generation prediction results,an intelligent collection mode of data collection nodes is constructed.The results show that the RMSE and MAE values of the SSAE-LSTM1 model are 1.65 and 1.70 lower than that of the multivariate Bi LSTM model without the SSAE feature learning layer,which proves that the feature extraction through SSAE can effectively improve the accuracy of photovoltaic power generation prediction.(2)Aiming at the problem that data transmission cannot be carried out in communication blind areas such as remote rural areas,a data communication mechanism based on Beidou short message is proposed.Combined with the characteristics of Beidou short message communication,the user data part of Beidou short message communication is designed to solve the problem of insufficient communication capacity;the sender request confirmation model and mechanism are proposed to enable the sender to transmit reliably and in different environments.Maximize transmission efficiency;propose a dynamic accumulation confirmation mechanism at the receiving end to ensure reliable communication in many-to-one communication scenarios,and verify the feasibility of the system mechanism through simulation experiments.In the field test,the transmission success rate of the sender changes dynamically with the environment so that the success rate of data transmission is increased to more than 99%.(3)Based on the above research,designed and developed an agricultural environmental data collection system,including a Web application system for agricultural environmental data and a We Chat applet system,with the collection,reporting,query,analysis and management of agricultural environmental data,etc.Function to provide a feasible solution for the development of smart agriculture.