CFD based heat transfer parameter identification of greenhouse and greenhouse climate prediction method

Predicting the greenhouse environment is important for crop yield and energy consumption projections. This study proposes a dynamic greenhouse temperature and humidity modeling method based on integrating CFD simulation and measured data. This method utilizes computational fluid dynamics to analyze the greenhouse's two key heat transfer characteristics. The effect of wind speed and temperature differences on the heat transfer coefficient of the cover is first analyzed under the condition that the inner thermal screen is folded. Secondly, the effect of temperature differences inside and outside the greenhouse, outdoor wind speed, sky temperature, and the internal thermal screen on the heat transfer coefficient between the greenhouse's indoor and outdoor environments is investigated. Based on the thermodynamic parameters of the greenhouse microclimate obtained through CFD simulation, a regression model of the greenhouse climate model parameters with respect to the influencing variables is built using a nonlinear identification algorithm. By analyzing the mechanism of greenhouse environmental changes, this work proposes a universal structure for the agricultural greenhouse climate model and uses identified heat transfer parameters to correct the universal mechanism model of the greenhouse climate. Considering complex environmental dynamics and unmodeled factors, this study proposes a weighted parameter correction method based on adaptive particle swarm optimization (APSO). Finally, the model was validated using climate data from Venlo-type greenhouses in Shanghai. The validation results indicate that the greenhouse climate model can accurately approximate the real greenhouse climate.