Prediction of the energy consumption is the most important topic for planning to build an energy power station. This energy power station can be non-renewable sources power plants or renewable power plants like wind and solar. Prediction of the energy consumption also figures out load modeling problem in new smart grid applications. In this study, energy consumption model is developed for temperature control of a greenhouse. Artificial Neural Network based modeling is advanced with temperature of inner, temperature of outer and temperature of soil. So, these temperatures are inputs in the ANN based model. In addition, the output of the ANN is energy demand that is strongly related with temperature data.