Umudike Journal of Engineering and Technology

In developing countries like Nigeria, where demand for power is increasing dramatically, it has become essential for electric power distribution companies to carry out projections regarding anticipated Electric Peak load demand in the future. In the planning and management of power generation, transmission and distribution systems, load forecasting is a key factor. This research presents the development of an Artificial Neural Network (ANN) based long-term load forecasting model for Aba, Abia State Nigeria. The recorded yearly load for 2012, 2013 and 2014 were obtained from (Enugu Electricity Distribution Company EEDC) and used in predicting the load for 2015. The Levenberg-Marquardt optimization technique which has one of the best learning rates was used as a back propagation algorithm for the Multilayer Feed Forward ANN model using MATLAB® R2008b ANN Toolbox. Several networks architectures were trained and simulated before arriving at the best Mean squared error performance of 1.14179e^-027 as depicted in Figure 6. It is also observed that from the Neural Network results, the peak load could reach 357.475MW by 2015, which is not far from the actual load demand peak of 353.32MW for the same year. The results have shown that the proposed technique is robust in forecasting future load demands for the yearly operational planning of power system distribution sub-stations in Aba, Nigeria.