Load forecasting: STLF using FFNN model with Adam Optimization Algorithm

Short-Term Load Forecasting (STLF) is the process of predicting the amount of electricity demand that will be required over a short period of time, ranging from a few hoursdays. The purpose of STLF is to help utility companies and power system operators to optimize the allocation and usage of energy resources, ensuring that enough electricity is generated and delivered to meet the expected demand at a given time. This involves analyses of factors that influences energy demand, such as weather conditions, time of day, weekdays, special events and other techniques to generate accurate and reliable predictions of future energy demand. Feed Forward Neural Network (FFNN) model is used for STLF in order to meliorate load accuracy. The data of three regions has been reckoned for training proposed model. Parameterstemperature, relative humidity, liquid precipitation and wind speed at a distance of 2m of every place is manipulated as training data set for presented model. Adam Optimization is applied for training model as it has robust learning rate, which makes this most pertinent algorithm to be applied for deep learning. This algorithm exhibits faster computational time and entails less tunning variables that embodies Adam Optimization as better algorithm and hence will improve accuracy outcomes of FFNN model