Knowledge of the variability of the solar resource in a geographical area of a photovoltaic energy production project is essential. At the time of decentralization, knowledge of data at the local level becomes critical for new projects of electrical autonomy based on photovoltaic energy in municipalities. Given that the density of the meteorological network in Togo is limited and only covers a small number of rural areas, the Adaptive Neuro-Fuzzy Inference Systems (ANFIS) method was used in this work to predict solar radiation across the entire territory, with the aim of providing a solar radiation data base for each municipality. The method employed exhibited high accuracy with the coefficient of determination (R2) higher than 98%, mean bias error (MBE) equal to -0.241 and root mean square error (RMSE) equal to 0.045355. Therefore, the data generated from the established models were used to realize solar irradiation Atlas with the Quantum Geographic Information System (QGIS) software to offer a geographical and spatial resource distribution. Therefore, hourly average irradiation maps of the annual and each of the twelve months have been prepared. The results show that Togo has a solar potential favorable for photovoltaic energy production. The sunniest months are January, February, March, April, October, November and December with the north of the country richer in solar potential than the south. [ABSTRACT FROM AUTHOR]