Short-Term Load Forecasting Using Time Pooling Deep Recurrent Neural Network

TL;DR

This paper proposes a novel Time Pooling Deep Recurrent Neural Network for short-term load forecasting, effectively handling high volatility data and improving accuracy over existing methods in smart grid applications.

Contribution

Introduces a time pooling strategy within a deep recurrent neural network to enhance load forecasting accuracy and reduce overfitting in volatile residential load data.

Findings

Outperforms existing algorithms in RMSE and MAE metrics

Effectively handles high volatility and data uncertainty

Reduces overfitting through data augmentation

Abstract

Integration of renewable energy sources and emerging loads like electric vehicles to smart grids brings more uncertainty to the distribution system management. Demand Side Management (DSM) is one of the approaches to reduce the uncertainty. Some applications like Nonintrusive Load Monitoring (NILM) can support DSM, however they require accurate forecasting on high resolution data. This is challenging when it comes to single loads like one residential household due to its high volatility. In this paper, we review some of the existing Deep Learning-based methods and present our solution using Time Pooling Deep Recurrent Neural Network. The proposed method augments data using time pooling strategy and can overcome overfitting problems and model uncertainties of data more efficiently. Simulation and implementation results show that our method outperforms the existing algorithms in terms of RMSE and MAE metrics.