It is of great importance to diversify electrical energy sources in recent times. In this study, the electrical energy that a house may need is provided by two different sources. The first of these is the city grid and the other is the battery that can be charged with solar energy. Transitions between two energy sources are carried out with the developed system. Thus, the energy costs of the home and / or office user are reduced and at the same time, energy estimates are made for future energy needs depending on the usage habits. Polynomial linear regression and LSTM methods were used for this estimation. Using the RMSE metric, we compared which method predicts with less error rate. For the nonlinear data set, LSTM performed more successfully, while for linear data such as electrical energy, the best estimation result was 0.99 with Polynomial Linear Regression.