Modeling And Forewarning Technology For Greenhouse Temperature And Humidity Systems

With the promotion of agricultural facilities and its supporting technology,our modern greenhouse showing large and wide features.Greenhouse is an independent special microclimate environment created by artificial work,and it is difficult to maintain ecological factors at crops breeding optimal interval.Greenhouse equipments control effect with inertia and lag,need a longer response time.In order to avoid extreme environmental conditions cause damage to the crops,important environmental factors modeling and short-term disaster forewarning can help adjust cultivation strategies timely,and provide a scientific basis to achieve “plant factory”.In this thesis,we use measured data inside and outside greenhouse as the research object,analysis of greenhouse microclimate model and its application in greenhouse environmental early warning.The main work has the following several aspects:(1)Greenhouse environment information acquisition.Choose stable and reliable equipments which can work for a long time,devices connected to the controller by building a unlimited access ZigBee network.Adopting the strategy of distributed monitoring,to transmit greenhouse data to the network using GSM/GPRS and to deliver data access remote terminal,realize the unattended environmental information collection.(2)Monitoring data preprocessing.Due to the measured data by the outside noise interference,using Pauta criterion to eliminate gross errors and wavelet denoising method smooth the data.A single set of sampled data is not representative,using batch estimation adaptive weighted fusion algorithm to handle multiple sets of data,providing a reliable modeling data samples.for subsequent analysis.(3)Factor selection and modeling method research.Through the analysis of the mechanism of greenhouse climate change process,we know many environmental factors that affect the model structure.Using Pearson correlation coefficient to determine which environmental factors required for modeling.Using system identification method based on measurement of input and output data to establish greenhouse temperature and humidity models.Akaike Information Criterion(AIC)is used to identify model structure,Recursive Least Square(RLS)is used to identify model parameters.(4)Environmental forewarning method study.According to the professional knowledge of agricultural expert to build crop disasterknowledge base and warning rule base,Using contain environment model,real-time data and knowledge base of forewarning method on crop disaster prevention and mitigation.The platform development of greenhouse ecological monitoring and forewarning system based on the LabVIEW.The simple interface and graphical menu easily accepted by the user.