Transmission System Load Forecasting Using Neural Network Approach, Case Study: Electricité Du Cambodge (EDC)

Transmission System Load forecasting is very significant in a developing country like Cambodia for an ideal decision in power system planning and controlling. However, the specific research on transmission load forecasting in Cambodia is still in small amounts and insufficient to meet the optimal planning for the rapid growth of the demand. This paper presents an attempt at a medium-term forecast for the transmission system load of the Electricité Du Cambodge (EDC) company located in Cambodia, using the Neural Network (NN) in comparison to the Linear Regression (LR) model. This study considered the monthly actual historical of transmission system load data from the year 2010 to 2019 which were recorded by EDC, and the input variables were calendar (date, months of the year, non-working days each month), and weather factors (relative humidity and wet bulb temperature). The forecasting performance of these two methods was evaluated by using the Mean Absolute Error (MAE) and the Mean Absolute Percentage Error (MAPE). Base on the performance result of the two models, the Neural Network (NN) model is more effective and the obtained result would be critically contributed to optimal planning for the Cambodia electricity transmission system network.