Remote Environmental Monitoring And Prediction System For Abalone Mushroom Greenhouse Based On Internet Of Things

With the improvement of people’s living standards,abalone mushroom,a rare edible fungus,has gradually been studied by many people,and some areas have been put into commercial production.Abalone mushroom,as a food ingredient,has a fresh and tender taste,containing various minerals,vitamin A,protein,amino acids,crude fat and other nutrients.These nutrients can reduce the accumulation of body fat,enhance muscle vitality,reduce cholesterol and blood lipids,soften and protect blood vessels,and prevent certain cardiovascular diseases.Abalone mushrooms also have the effects of nourishing the spleen and stomach,dispelling wind and dispelling cold,and can alleviate symptoms such as waist and leg pain and numbness in hands and feet.The cultivation cost of abalone mushroom is low and the income is high,but the mushroom needs to have a relatively strict growth environment,which requires real-time monitoring and adjustment of the cultivation environment.Based on the above requirements,this thesis designs and implements a remote environmental monitoring and prediction system for abalone mushroom greenhouses based on the Internet of Things.According to the characteristics of the Internet of Things,the system is divided into three parts:a lower computer,a remote monitoring cloud platform,and a Web client.The lower computer design consists of a parameter acquisition module with STM32F103 as the main control chip,a control module with STM32F407 as the main control chip,and a power module.The power module supplies power to all modules.The parameter collection module is responsible for collecting environmental parameters and transmitting the collected data to the control module through RS485 communication.The control module then processes the data and transmits it to the remote monitoring cloud platform through Ethernet communication.The cloud platform is responsible for data storage and interaction.The system’s operating environment and My SQL database are deployed on it.The upper computer is composed of a cloud platform and a client,and is developed using Java Web technology using a Browser/Server architecture model.The client sends instructions to the cloud platform at the set frequency to request environmental data,and displays the returned data visualization visually on the web page.It can also remotely control the equipment automatically or manually when environmental parameters are abnormal,and send SMS alarms to generate alarm logs.It can view historical data to generate daily curves of each environmental parameter,and can predict to generate environmental parameter prediction curves.Based on the prediction image and evaluation indicators RMSE（Root Mean Square Error）,MAE（Mean Absolute Error）,MAPE（Mean Absolute Percentage Error）,R^{2（）R Squared},this thesis compares the advantages and disadvantages of several kinds of algorithms:BP（Back Propagation）neural network,LSTM（Long Short Term Memory）neural network,EEMD-BP（Ensemble Empirical Mode Decomposition-Back Propagation）algorithm and EEMD-LSTM（Ensemble Empirical Mode Decomposition-Long Short Term Memory）algorithm,and concludes that the EEMD-LSTM prediction model has the best effect.Providing the function of environmental parameter prediction can give managers enough time to take corresponding measures to avoid unnecessary losses.The results of many tests for a long time show that the system can monitor the environmental parameters of the greenhouse remotely and in real time,regulate and control the environmental equipment under abnormal conditions in time or even in advance,and the data collection,transmission,environmental regulation,early warning and other aspects are consistent with the expected results.The stable operation of various functions has certain practical significance for the cultivation of abalone mushroom to improve the mushroom yield and increase income.