In Brazil, due to government incentives given in the area of renewable energy, a growing perspective in the number of micro and mini distributed generation (MMGD) installations is postulated, being the solar source highlighted in the country. Thus, the increase in the insertion of intermittent sources promotes significant changes in the behavior of the hourly load curve, which can directly affect the operation and planning of the electrical network. To meet the new panoramas provided by the Brazilian electricity system, this dissertation proposes a new methodology to account for distributed photovoltaic generation for the hours that make up the day. Using the Double Seasonal Holt-Winters model, load and demand forecasts are made for the National Interconnected System and the subsystems that integrate it, considering, in particular, the impact caused by the connection of these solar photovoltaic MMGD systems in the distribution network. For the forecasts, the 24-hour time horizon is used, in hourly intervals, carried out for the first week of 2020. The results indicate that the proposed methodology for the creation of distributed photovoltaic generation series is valid, as it is observed a decrease in the forecast errors for the demand series, constituted by the amount of the distributed generation added to the load. The MAPE values analyzed in this work do not exceed 10 percent for typical days, except holidays, indicating that the presented method is an efficient resource.