OPTIMAL DISTRIBUTION OF ELECTRICAL ENERGY IN THE CITY OF GOMA

This work explores the use of neural networks to predict electricity demand and production and to optimize energy distribution in Goma. It aims to develop ANNs capable of anticipating production at the RUZIZI MATEBE RWANGUBA NURU and NELSAP power plants and to design a model optimizing energy distribution. Following a literature review and the collection of local and online data (Kaggle) an ANN model was trained. The results show demand prediction with an error of 1.43% and distribution optimization with an accuracy of 90%. The study reveals that NURU and MATEBE are crucial during peak periods albeit costly and generating high losses. This approach significantly improves resource management compared with conventional methods.