The diversified and efficient use of sustainable energy resources is becoming more crucial day-by-day in today's energy-dependent world. In response to such a situation, alternative energy sources should be evaluated, especially on a regional basis. The recent pandemic and its effects show that the effective and widespread use of green energy sources should be increased, because air pollution causes an increased rate of COVID19 virus transmission. It is necessary to re-plan all sustainable energy resources on a regional basis because of the negative environmental impacts of burning fossil energy. On a global scale, there is an initiative to reduce carbon dioxide (CO2) emissions related to renewable energy sources by 70% by 2050. In many applications, efficiency analysis and new technologies are among the primary topics. As the rate of urbanization rises, there is a great need for integrated new energy technologies that control local energy efficiency standards for cities, or even buildings. Urban planning is crucial to reducing energy use in buildings. Turkey aims to install 34 gigawatts (GW) of hydroelectric capacity, 20GW of wind energy, 5GW of solar energy, and 1GW power from geothermal or biomass-based on strategic green energy plans. Turkey will provide about 30 % of its total energy needs via renewable energy sources based on this plan by 2023. As is known, one of the most important sustainable energy sources is solar power derived from solar radiation. Solar power has a characteristic pattern compared to other sustainable energy sources, not only in seasonal but also in daily terms. The geographical requirements, vast flat terrain, and high irradiation levels of solar power plants make it ideal for promoting economic growth in the Central Anatolian inner regions of technological investments. In this study, we present estimations based on model results for the monthly variation of solar energy potential, efficiency. Study areas are in the vicinity of Ankara (Camlidere and Kecioren) in Central Anatolia. To forecast solar energy potential, hourly solar radiation (watt/m2) was processed with five years of data (from 2014 to 2018). Deep learning methodologies were applied to solar radiation data to build up future solar radiation data scenarios in 5-year forecast through the end of 2024. A 3-layer artificial neural network (ANN) consisting of 128-neurons with Rectified Linear Unit (ReLU) activation was applied to the solar data. An LSTM layer using 64 neurons with ReLU activation functions for each neuron, was applied to build the hidden layer up. The output layer was built on the hidden dense layer using 2-neurons with a linear activation function. Solar power efficiency and performance of ANN models in the annual basin are presented in two study areas. The model's performance increases by up to 98% to estimate solar radiation and sunshine duration. There is sufficient evidence based on the results to invest in renewable energy sources using solar energy converting systems at both study areas.