Modeling of greenhouse climate using evolutionary algorithms.

摘要

Because the Mexican greenhouse industry is growing rapidly, it is important to increase the knowledge on the indoor environmental conditions of greenhouses. In the current research, a first principle mathematical model was developed to account for the behaviour of air temperature, humidity, and soil temperature under greenhouse conditions in the central region of Mexico. The model structure contains three state variables: air temperature, air humidity, and upper-layer soil temperature. Several parameters like wind effect coefficients were also estimated. However, because of the nonlinear nature of the model, convergence to local minimum was observed as local search methods, such as nonlinear least squares was applied for model calibration. Therefore, efficient evolutionary algorithms named differential evolution (DEAs) algorithms were used. Convergence of DEAs was more consistent to an apparently global optimum. Furthermore, a better agreement between predicted and measured values for the three state variables using evolutionary algorithms was observed. Therefore, DEAs can be an alternative in estimating the parameters in complex models of greenhouse climate.