Predicting greenhouse daily light integral using photosynthetic photon flux density and finite difference method

Abstract

In the state-of-the-art agricultural technology using greenhouses, it is useful to estimate a dynamic model of internal environmental changes according to external environmental changes. In this article, we propose a method to predict the change in the daily light integral inside the greenhouse according to the change in the daily light integral outside the greenhouse. Since daily light integral corresponds to the integral value of photosynthetic photon flux density, it is necessary to obtain a model of change in photosynthetic photon flux density inside the greenhouse according to the change in photosynthetic photon flux density outside the greenhouse to improve the prediction accuracy. The greenhouse was considered a nonlinear first-order system with a time lag for the photosynthetic photon flux density, and the system parameters were identified using the recursive least-squares method. Here, the photosynthetic photon flux density data measured outside the greenhouse is used as the input to the system, and the photosynthetic photon flux density data measured inside the greenhouse is used as the output of the system. Next, we obtain the daily light integral prediction model. The daily light integral prediction model is estimated from the photosynthetic photon flux density model using the finite difference method. The performance of the proposed method has been evaluated through extensive experimental studies. In Korea, since the weather changes in summer and winter are great, it must compare the daily light integral prediction performance in summer and winter to more accurately verify the proposed method. In this study, the experimental results are presented, respectively, for summer in August 2021 and winter in January 2022. The error performance of the human empirical prediction method and the proposed data-driven prediction method are compared for two seasons, and the robustness of the rapid weather changes of the proposed method is discussed.