Soil Moisture Detection And Irrigation Decision Support Based On The Internet Of Things

China has always been a water-poor country.While the annual precipitation is low,the distribution of precipitation is also uneven.Insufficient soil moisture will cause drought in the cultivation process of crops and affect the normal growth and development of crops,so timely It is very important to obtain soil moisture content information and use it as a basis for controlling soil moisture content to ensure the moisture needed for crop growth.This paper studies the cloud platform technology of the Internet of Things and agricultural sensor technology and combines it with the mobile Internet,and designs a remote transmission system that can upload soil moisture information to a cloud server,realizes remote monitoring of soil moisture information,and uses crop coefficients and Penman-Monteith formula combined with meteorological data to calculate the actual water demand of crops,combined with BP neural network to predict and set fuzzy irrigation rules to achieve irrigation decision support.The main research contents are as follows:(1)Using sensors,STM32 microcontroller,wireless communication module,and cloud server to complete the soil moisture detection system,the collection of soil moisture content,environmental temperature and humidity,light and wind speed information is realized,and the corresponding software part is designed to upload the information to the cloud server.(2)Using penman Monteith formula and crop coefficient to calculate crop water demand,combined with BP neural network modeling prediction,the prediction results of the initial growth period R2 is 0.9543,the middle growth period R2 is 0.97521,the mature period R2 is 0.98209,which shows that the three models have a good performance,and then set irrigation rules to realize the theoretical support of irrigation decision.(3)Compared with the current domestic mainstream cloud platform providers of Internet of things,Alibaba cloud was selected to design the cloud server,and the cloud server was designed and deployed.Combined with the web visualization development function in the cloud server,the sensor information was displayed,and the remote monitoring function was completed,and the development process was introduced.(4)After the completion of all parts of the design and construction,completed the information upload,display,proved the normal function of the system.