Estimating production in a geothermal energy healed greenhouse with artificial neural networks

As in every field, reliable and accurate estimations for the future are needed in agricultural applications. Since artificial neural networks are used effectively in solving estimation problems and provide very reliable results, the use of this technique has become increasingly widespread.In this context, in the thesis study, production estimation was carried out in a greenhouse heated with geothermal energy with an artificial neural network. A new artificial neural network model was established by using the data of two production seasons from the greenhouse where the study was carried out. Levenberg – Marquardt algorithm was used in the developed artificial neural network model. The actual production values taken from the greenhouse and the estimation results produced by the artificial neural network model were compared. In this way, it was seen that the developed model made an accurate estimation of 99.43% in the total production amount.In addition, comparisons were made by looking at the values in other commonly used error measures. As a result, according to the findings, it was seen that the artificial neural network model showed high performance in estimating the total production amount. Therefore, it has been concluded that artificial neural networks can be a model that can enable manufacturing enterprises to predict the situations they may encounter in the future with the current data and to make highly accurate decisions.