PREDICTION OF INDOOR TEMPERATURE IN A GREENHOUSE: SIIRT SAMPLE

Abstract

Greenhouses are agricultural structures which allow indoor conditions to be controlled. Food demand is increasing as the population increases. Therefore, creating new areas for food production and making perpetual agricultural production without interruption is a key-stone necessity to meet demands. Increasing the crop production could only be possible with constant cultivation period. Green- houses are the capital ships of feeding the population and fighting poverty. Due to the climate changes and increasing population, greenhouses will gain more and more significance in the years to come. However, greenhouses will cause harms instead of benefits if they are applied in wrong climatic conditions. In this study, in a greenhouse having floor area of 11220 m2, indoor and outdoor temperatures are quantified for two years, after modelling and simulating the energy efficiency, indoor temperature values are estimated by artificial neural networks. This study shows that artificial neural networks could accurately estimate the indoor temperature of greenhouses and relative hu- midity 6 hours in advance, and the temperature could be estimated 3 days in advance.