Modeling and Simulation of Temperature and Relative Humidity Inside a Growth Chamber

Abstract

Modeling and simulation of internal variables such as temperature and relative humidity are relevant for designing future climate control systems. In this paper, a mathematical model is proposed to predict the internal variables temperature and relative humidity (RH) of a growth chamber (GCH). Both variables are incorporated in a set of first-order differential equations, considering an energy-mass balance. The results of the model are compared and assessed in terms of the coefficients of determination (R 2 ) and the root mean squared error (RMSE). The R 2 and RMSE computed were R 2 = 0.96, R 2 = 0.94, RMSE = 0.98 • C, and RMSE = 1.08 • C, respectively, for the temperature during two consecutive weeks; and R 2 = 0.83, R 2 = 0.81, RMSE = 5.45%RH, and RMSE = 5.48%RH, respectively, for the relative humidity during the same period. Thanks to the passive systems used to control internal conditions, the growth chamber gives average differences between inside and outside of +0.34 • C for temperature, and +15.7%RH for humidity without any climate control system. Operating, the GCH proposed in this paper produces 3.5 kg of wet hydroponic green forage (HGF) for each kilogram of seed (corn or barley) harvested on average. Energies 2019, 12, 4056 2 of 22techniques are analyzed to maintain the following critical variables: Temperature, humidity, lighting, and carbon dioxide concentration [1,2]. Other research projects predict the behavior of one or more variables and provide an appropriate response to keep these variables within the desired limits [3][4][5]. Most of the research on modeling and simulations has mainly been developed for greenhouses [6][7][8][9][10][11] to predict the internal environmental conditions of enclosures utilized for cultivation and plant growth. Several studies from the perspective of different environments of models have also been pursued, with the objective of knowing the behavior of the internal variables, such as temperature and relative humidity, that will serve as support to later define an adequate strategy for controlling the growth conditions within an enclosure. Although the type of crop is not defined explicitly, its influence is implicit, given its inherent condition in the models. These studies have implemented different approaches to develop these models, such as neural networks, genetic algorithms, and neuro-fuzzy models [12][13][14][15][16], where the object of study centers on natural ventilation, forced ventilation, cooling evaporators, or systems that integrate calefaction [17][18][19][20], to mention a few of them. On the other hand, there are alternative studies not developed for greenhouse environments, and proposed by some authors to model and simulate temperature and humidity [21][22][23][24][25].