Using neural network for long term peak load forecasting in Vientiane municipality

Long term peak load forecasting (LTPLF) plays an important role in electrical power systems in term of policy planning and budget allocation. The planning of power system expansion planning starts with the forecast of anticipated load requirement. Accurate load forecast can be helpful in developing a power supply strategy, and development plan, especially for developing countries where the demand is increased with dynamic and high growth rate. This paper is proposed a peak load forecast model using artificial neural networks (ANN) to employs multilayer perceptron (MLP) with backpropagation (BP) learning algorithm. This method can be used to forecast the monthly peak load in a year from one month to three years. The factors that reflect weather and economic variation are selected based on correlation coefficients. In addition, a new training technique is introduced. A case study is performed by using the proposed method of peak load data of Vientiane municipality system of Lao PDR to demonstrate accuracy of the proposed method. The forecast model is shown to be simple and highly accurate. The mean absolute percentage error (MAPE) of the model is 0.19-4.09%.