14-11-2014, 10:21 AM
Abstracts: Load Forecasting is considered to be the most prominent part of Power System management. Accurate Load Forecasting can be very helpful in making Unit Commitment problems solutions, reducing spinning reserve capacity and maintenance scheduling of power system. Short term load forecasting means a day ahead forecasting considering various parameters. For this purpose Artificial Neural Network (ANN) method is used. Here Two different networks are created i.e. (1) Regular Days Network (Monday to Friday) (2) Weekend Days Network (Saturday and Sunday) due to their different load patterns. For the training of these networks the full month data of February 2014 was taken from GUJARAT ENERGY TRANSMISSION CORPORATION (GETCO). The inputs used were the hourly load demand for the full day (24 hours) for the state and the daily temperature and humidity major cities. The outputs obtained were the predicted hourly load demand for the next day. The neural network used has 3 layers: an input, a hidden, and an output layer. The input contains 72 neuron while training of network using different number of neurons in hidden layer was tested and was settled down at 105 neurons for regular days and 25 neurons for weekend days. The forecasting of the load using ANN gives a high degree of accuracy and a Mean Absolute Percentage Error (MAPE) of 3.45% was observed.