Intelligent state estimation for future Nigerian smart grids Using ANN

The increasing integration of renewable energy, distributed generation, and mini-grids in Nigeria highlights the need for accurate and real-time state estimation in distribution networks. Conventional Weighted Least Squares (WLS) estimators, while widely used, face challenges in modern smart grids due to computational complexity, sensitivity to noise, and iterative convergence issues. This study proposes an adaptive Artificial Neural Network (ANN)-based state estimation framework tailored for emerging Nigerian distribution networks. The model is evaluated using the IEEE 33-bus test system, serving as a representative benchmark for Nigerian feeders. Results show that the ANN estimator significantly improves accuracy and efficiency, reducing voltage RMSE from 0.0075 p.u. to 0.0032 p.u. and angle RMSE from 0.85° to 0.38°, while achieving nearly six times faster computation than WLS. These findings demonstrate the ANN framework’s potential as a scalable, intelligent tool to support reliable and resilient smart grid operations in Nigeria.