Deterioration of Electrical Load Forecasting Models in a Smart Grid Environment

Abstract

Smart Grid (S.G.) is a digitally enabled power grid with an automatic capability to control electricity and information between utility and consumer. S.G. data streams are heterogenous and possess a dynamic environment, whereas the existing machine learning methods are static and stand obsolete in such environments. Since these models cannot handle variations posed by S.G. and utilities with different generation modalities (D.G.M.), a model with adaptive features must comply with the requirements and fulfill the demand for new data, features, and modality. In this study, we considered two open sources and one real-world dataset and observed the behavior of ARIMA, ANN, and LSTM concerning changes in input parameters. It was found that no model observed the change in input parameters until it was manually introduced. It was observed that considered models experienced performance degradation and deterioration from 5 to 15% in terms of accuracy relating to parameter change. Therefore, to improve the model accuracy and adapt the parametric variations, which are dynamic in nature and evident in S.G. and D.G.M. environments. The study has proposed a novel adaptive framework to overcome the existing limitations in electrical load forecasting models.

Machine Learning-Based Load Forecasting for Nanogrid Peak Load Cost Reduction

“…Furthermore, load data have been decomposed into observed, trend, and seasonality. To show load variations, the decomposed data from 1 January 2016 to 15 January 2016 have been plotted in Figure 6 as an example, in which it is evident that the load does not have a daily trend or specific pattern on a day-to-day basis; unlike gird or microgrid [47,54]. In addition, seasonality can be observed on a day-to-day basis, and it oscillates from −0.5 kW to 0.5 kW, which also shows that load has variability.…”

Real-Time Load Forecasting and Adaptive Control in Smart Grids Using a Hybrid Neuro-Fuzzy Approach

“…Studying the use of REPS-based FIOS for load forecasting, with an emphasis on cutting-edge fuzzy logic methods, was the goal of Forecasting through Estimated Convergence of REPS-based Fuzzy Inference Systems in Smart Grids [14]. More effective energy management and enhanced grid performance were two outcomes of a suggested advanced forecasting system that used an adaptive neuro-fuzzy inference system and a genetic algorithm to anticipate electrical demand [15].…”

A Deep Learning Approach to Real-time Electricity Load Forecasting

“…However, decisions concerning prosumers within the smart energy grids will require a short timeline, namely, minutes to days. In addition, medium-term forecasts, weeks to months ahead, are crucial in power systems scheduling [3][4], while long term forecasts, monthly/yearly predictions, support grid maintenance planning.…”