Application of Recurrent Neural Network Model in the Analysis of Electricity Load Demand in Ashanti Region of Ghana

In order to supply electric energy to the customer in a secure and economic manner, an electric company faces many economical and technical challenges in operation. These challenges include scheduling, load flow analysis, planning and control of electric energy system. To address this problem, accurate models for electric power load forecasting are essential to the operation and planning of utility companies. The study used a data from the Ashanti East and Ashanti West Regions of the Electricity Company of Ghana. The data is a daily peak load for the whole year of 2014, 2015 and the first month of 2016 with a total of 752 samples. In this study three Recurrent Neural Network(RNN) with layered architecture models (RNN-1-10-1, RNN-1-15-1 and RNN-1-20-1) were developed to forecast the peak electricity load. The test results showed that the best model which forecast the peak load well is RNN-1-10-1 with minimum RMSE value.