Research On Solar Greenhouse Modeling And Intelligent Control Strategy

The per capita arable land area is far lower than the world average in China.Especially in the northern regions,the winter is cold,and the annual growth cycle of crops is short,resulting in insufficient supply of major agricultural products such as vegetables,fruits and fruits in winter.With the advancement of technology,greenhouse planting can provide a comfortable growing environment for crops throughout the year,so that crops that were originally produced under certain seasonal conditions can be produced off-season,increasing the supply of agricultural products,and the quality is correspondingly Improvement.At present,China has begun to vigorously promote greenhouse cultivation,becoming the country with the largest greenhouse planting area in the world.According to the growth needs of plants,it is particularly important to control the environmental factors inside the greenhouse,especially the temperature and humidity.In order to achieve the control of the temperature and humidity inside the greenhouse,it is necessary to analyze the law of temperature and humidity change in the greenhouse,that is,the temperature and humidity mechanism,and to determine the factors affecting the temperature and humidity changes for greenhouse modeling.A precise mathematical model of temperature and humidity at room temperature is a prerequisite for precise control of intelligent greenhouses.This paper first studies the identification method of greenhouse model.According to the characteristics of multi-parameter,strong coupling,highly nonlinear and time-varying solar greenhouse,it can be identified by using neural network,or the algorithm can be used to identify the unknown parameters of the mechanism model according to the temperature and humidity mechanism,and obtain a relatively accurate mechanism model.Different methods can be used to obtain different models in different growth stages of crops to control the growth environment of greenhouse crops.In the crop planting period and seedling stage,the neural network identification model is adopted;the mechanism model after algorithm identification is adopted in the crop growth period.Then,by consulting the research results of domestic and foreign scholars,the model of temperature and humidity conforming to the greenhouse in northern China is summarized.The influence factors affecting the temperature and humidity inside the greenhouse are obtained from the mechanism model,and they are used as the input of the neural network model in Matlab/Simulink.The particle swarm optimization(PSO)algorithm,Levenberg-Marquardt(LM)algorithm and PSO-LM algorithm are used to optimize the RBF neural network weights and thresholds.The comparative analysis of the prediction results shows that the PSO-LM-RBF model can be used to construct The temperature and humidity model.It has the lowest prediction error and the fastestspeed,and can provide guidance for controlling the temperature and humidity in the greenhouse during crop planting and seedling stage.In the mechanism model modeling,the temperature and humidity mechanism was established according to the greenhouse mechanism,and the ratio coefficient of solar energy absorbed by plants related to crop photosynthesis and the heat transfer coefficient of leaves related to crop transpiration in the mechanism model were identified by the firefly algorithm,and finally a relatively accurate greenhouse mechanism model was obtained to control the temperature and humidity in the greenhouse during the crop growth period.Finally,the control system corresponding to the neural network model and the mechanism model is designed separately,and the control system is reasonably selected during different growth stages of the crop,so as to realize the intelligent control of the temperature and humidity inside the greenhouse.