Abstract: Accurate prediction of electricity consumption is essential for the efficient functioning of power grids and the successful administration of energy markets in the field of energy planning. There are several benefits such as system efficiency, dependability, safety, and stability with the proper forecasting of load demand. To ensure the proper operation of the microgrid energy management system, it is essential to predict the overall load demand with precision and consistency. A bidirectional long short-term memory network (Bi-LSTM) model is considered in this paper for the electrical load forecasting of microgrids. The proposed model is verified by using the standard available load forecasting data set of Panama City. Furthermore, the approach is compared with LSTM load forecasting method. The various performance metrics such as MAE, RAE, RSE, R2, RMSE, and NRMSE are employed in this paper to evaluate the accuracy of the proposed prediction model. In this work, the Bi-LSTM method got the better error reduction values of 17.8% in MAE, 1.70% in RMSE, 2.37% in NRMSE, 3.5% in RSE, 19% in RAE and 1.2% improvement in R2 when compared to the LSTM model. The proposed load prediction model is helpful in estimating the future power shortages of the microgrid community which leads to improve the system efficiency and reliability.
Venue: Raipur
Institute: National Institute of Technology Raipur
Home »