One day ahead daily peak hour load forecasting by using invasive weed optimization learning algorithm based Artificial Neural Network

Load forecasting has been essential part of an efficient power system planning and operation. It is a pre-condition to economic dispatch of electrical power and improves the accuracy beside ascertain reliable operation of a power system. Normally the electrical energy demand is mostly dependent on various independent variables such as day, time, temperature, weather and holidays in a week. The load forecasting sensibility is a key to make sure the electrical energy supply to customers without harm in economic aspect of power system operation. In this project, an Artificial Neural Network (ANN) trained by the Invasive Weed Optimization (IWO) learning algorithm is proposed for short term load forecasting (STLF) model. By using 'seen' and 'unseen' of electrical energy demand data were used to test the performance of the proposed algorithm. Based on result obtained, it shows that IWO learning algorithm is capable to produce accurate prediction load demand. Hence, this indicates that Invasive Weed Optimization could be implemented as a new learning algorithm for an Artificial Neural Network.