Greenhouse Environment Optimization Method Based On CFD And Model Reduction Technology

As a semi-closed small ecosystem,on the one hand,agricultural greenhouse can reduce the dependence on external conditions,on the other hand,the weak self-regulation ability determines that it needs manual intervention to improve its internal environment and improve crop yield and quality.The traditional greenhouse environment regulation strategy based on sensor feedback cannot accurately reflect the influence of the spatial distribution of environmental parameters inside the greenhouse,and there are problems such as low regulation precision and poor operating energy efficiency.This paper constructs a computational fluid dynamics(CFD)model of the greenhouse environment,introduces model reduction methods,multi-objective optimization algorithms and other tools to improve the greenhouse crop environment and reduce the operating energy consumption of greenhouse-related equipment.Conduct research on operation optimization and design optimization of key parameters of the greenhouse system.The main research contents include:(1)According to the actual situation of greenhouse,the greenhouse CFD model is established.CFD software is used to simulate the internal environmental system of greenhouse;Comprehensively analyze the distribution law of environmental factors such as indoor temperature field,air flow field and carbon dioxide concentration field;The results of numerical simulation are compared with the experimental values to verify the accuracy of the established model and provide a model basis for the subsequent optimization of greenhouse environmental parameters.(2)A combined optimization framework of greenhouse environment based on CFD and model reduction is designed.The greenhouse wet curtain temperature and fan speed are set as optimization variables,and the temperature field and carbon dioxide concentration field in the crop area are taken as optimization objectives,taking into account the energy consumption generated by fan operation.Firstly,C ++ language is used to realize the multi-objective optimization algorithm,and the interface file is written to realize the interaction between CFD calculation and optimization algorithm.Secondly,based on the preliminary optimization solution obtained by the above CFD interactive method,the parameters are accurately optimized based on the reduced order model.This method uses the intrinsic orthogonal decomposition(POD)model reduction technology to establish the low-order subspace corresponding to the original CFD parameter space,and uses multidimensional interpolation to quickly solve the environmental response in each iteration of the optimization algorithm,so as to realize the rapid and accurate optimization of environmental parameters.It has the advantages of low time consumption and low calculation cost.(3)Under the platform of ANSYS Workbench,the greenhouse structure optimization design is carried out based on the response surface experimental design analysis method.Taking the location of the fan as the design variable and the indoor temperature as the optimization objective,the prediction regression model between the design variable and the response value is established and analyzed;Using the multi-objective optimization algorithm(MOGA)and the response surface model,the multi-objective optimization results of the fan position are obtained.The best candidate design points given by the algorithm are simulated by CFD.The results show that the temperature field of a section corresponding to the optimized fan position is about 0.4℃ lower than the average temperature of the original model,which verifies the effect of structural optimization design.