LONG-TERM LOAD FORECASTING USING ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM: A CASE STUDY OF 330KV MANDO TRANSMISSION STATION, KADUNA, NIGERIA

Long-Term Load Forecasting Using Adaptive Neuro-Fuzzy Inference System: A Case Study of 330kV Mando Transmission Station, Kaduna, Nigeria is a research work which is aimed at predicting future energy demand, in Megawatt (MW), with the adaptive neuro-fuzzy inference system (ANFIS) method. Electric load forecasting (ELF) is an informed use of previous and current load demands in predicting load demand to come with high measure of reliability and accuracy. Load forecasting (LF), is a requirement for generation and or transmission expansion. The ANFIS method is a hybrid Artificial Intelligence (AI) method which consists of Artificial Neural Network, Fuzzy Logic and Fuzzy Inference System to develop a model which is intelligent in decision-making. Time series load demand data (in MW) for 2022, 2023 and 2024 were collected from the Mando Transmission Station, analyzed and collated into Peak and Average daily load demand. These 3 years load data were daily, half-hourly load demand data up to 56,560 data set and therefore used for the long-term electric load forecasting. With the use of feed forward back propagation training algorithm, in MATLAB/Simulink, these data were trained, tested and forecasting of load demand for 11 years to come was done: 2025 to 2035. Assessment of the performance accuracy of the data training was done with Root Mean Square Error (RMSE) assessment metric. The maximum prediction error for the Peak load demand was 0.07% while that of the Average load demand was 0.08%. This shows successful training and high-performance accuracy of the ANFIS model. Forecast made for the 11 years horizon showed the maximum peak daily load demand will be 592.7MW and will occur on 1st of December, 2031 while the corresponding maximum average daily load demand will be 262MW and will occur on 11th of November, 2031. Furthermore, the results show the peak daily load demand will rise by 69% in eleven years to come while the average daily load demand will rise by 81% in the same period.