Long term load forecasting for Nigeria's electric power grid using ann and fuzzy logic models

This paper presents long term electric load forecasts for the Nigerian Transmission grid using the concepts of Fuzzy Logic (FL) and Artificial Neural Network (ANN), and forecasting accuracy by comparing the outcomes of these two approaches. The two techniques were developed and trained using econometric data such as gross domestic product and population. Data of monthly electricity supply of the 28-Bus 330kV grid from 2000 to 2014 were obtained and applied to create ANN and FL forecasting algorithm. The forecasting accuracy of these models were evaluated using absolute and mean absolute percentage errors. It was concluded that in long term load forecast, ANN and Fuzzy logic are powerful tools for load forecasting with minimum error rate. The obtained forecasts for Nigeria's 330kV grid in 15 years, using these techniques, were 7853MW for FL and 8189MW for ANN respectively. These values are applicable for transmission expansion planning on a long term.