ARTIFICIAL NEURAL NETWORK ALGORITHM BASED SHORT-TERM LOAD FORECASTING FOR MEDIUM VOLTAGE NETWORKS
Loading...
Date
2021
Journal Title
Journal ISSN
Volume Title
Publisher
ARPN Journal of Engineering and Applied Sciences
Abstract
Electrical energy is generally known that it cannot be stored. Therefore, it is generated whenever there is need or demand for it. Thus, it is imperative for the power utility companies that the load on their systems should be estimated in advance while such estimation of load in advance is referred to as load forecasting. The forecasting could be Short term, Medium term and Long term depending on the certain parameters in consideration. Short term load forecasting method usually has period ranging from one hour to one week. It often assists in approximating load flow and to make decisions that can intercept overloading. Also, Short term forecasting provides obligatory information for the system management of daily operations and unit commitment. This paper presents an Artificial Neural Network-based model for Short-Term Electricity Load Forecasting. The performance of the model is evaluated by applying the hourly load data of a leading power utility company in Nigeria to predict the required load of the next day in advance. These hourly load data were obtained from two number 33KV feeders; namely the Government house and Sabo-Oke. Also, the data were normalized
and then loaded into the model. The model was trained in MATLAB R2020a neural network Simulink environment. The simulation results show a good prediction accuracy for the two domains.