Abstarct: According to the necessary prediction of building energy consumption, this thesis considers the external and internal disturbances and the artificial neural network is used to predict the electric energy consumption in a house, Tyler, Texas, the USA in 72 days of summer. Furthermore, in order to improve its performance, genetic algorithm is used to find the optimal structure of the neural network which includes the numbers of neurons and the active functions. The neural network provides satisfactory results in predicting consumption energy which the results regarding to the neural network with Optimized structure is also improved significationly. Finally, some approaches are suggested about energy management to decrease the effect of the external disturbances and the heat pump energy consumption the building.