Research On Multi Factor Association Intelligent Control Modular Greenhouse

The greenhouse can provide relatively closed and controllable environment for agricultural production.It can accurately control the input output rate,develop the potential productivity of the land and improve the efficiency of agricultural production through the application of new technologies,new processes and new management methods.In recent years,the scale of greenhouse and related technologies of it have developed rapidly,but the development of the greenhouse standardization is slow,which hinders the upgrading and popularization of the greenhouse structure technology and the production environment control technology.In view of the above problems,this paper combines the greenhouse structure technology and the greenhouse intelligent control technology to realize the serialization and modularization of the greenhouse based on the general intelligent control model.It lays the foundation for the research of greenhouse standardization to meet the needs of the coordinated development of structural technology and intelligent management technology,and is of great significance for promoting the development of facilities agricultural technology.The main research work in this paper was as follows:(1)Building up a greenhouse classification modelIn this paper,a greenhouse classification model is established based on hierarchical clustering algorithm.The classification and clustering of greenhouse configuration can be realized by selecting feature sets reasonably.The results show that the model can effectively classify the greenhouse configuration samples,and the clustering results are relatively stable when the threshold is within 0.2-0.5.But when the threshold is less than 0.2,the clustering is too much,when the threshold is greater than 0.4,the result is over clustering,such as clustering the solar greenhouse and arch shed with distinct structure.(2)The serialization and modular design of greenhouseBy analyzing and comparing the mechanical and technological properties of structural materials,a general structural material used in intelligent greenhouse is established.Based on this,a series of greenhouse parameters and spectral series were constructed.In order to realize serialization,the thought of greenhouse modular design and the principles and methods of greenhouse module division are put forward.Taking solar greenhouse as an object,the modular implementation mode is studied,and the interface rules of modules are established.(3)Research on micro greenhouse environment modelIn view of the internal structure and plant cultivation of the greenhouse,a micro greenhouse CFD model based on the laminar flow model and the multi space medium model was established to study the heat model of the greenhouse.The research shows that by reasonably controlling the position and section parameters of the vent,the ventilation of the greenhouse can be realized by using the artificial light source to dissipate heat.When the artificial light source is 15 W,the ideal cross section of the upper vent is 45 cm×1.5 cm.The internal temperature balance can not be realized by natural convection under the heat preservation mode of bothside low air vent or no air vent.The auxiliary convection or increasing temperature of the circulation equipment should be added.(4)A multi factor correlation greenhouse temperature prediction model based on BP neural network is studied.Taking the temperature of multi span greenhouse as the research object,a BP neural network model is established.The model sets 6 input layer neurons:greenhouse carbon dioxide,light intensity,air temperature,air humidity,soil moisture,soil temperature,an output layer neuron:air temperature.The single hidden layer has the best prediction.accuracy when tested with 13 neurons.The BP neural network model can better express the nonlinear relationship between the temperature and the main control factors.The mean square deviation between the predicted results and the measured values is less than 0.23%.The correlation coefficient of the test sample between the network output value and the target value of the network is above 0.95?(5)Technology implementation and verification of greenhouse modular and general control modelsThe technical requirements and technical limitations of environment for greenhouse were discussed.The combination and content adjustment of module technology make the performance of greenhouse modular technology conform to the actual use needs.The greenhouse temperature prediction model based on BP neural network is systematically implemented and verified.The results show that the prediction accuracy of the prediction model can meet the requirements of the greenhouse temperature prediction accuracy.