Multi-region electricity demand prediction with ensemble deep neural networks

Abstract

Electricity consumption prediction plays a vital role in intelligent energy management systems, and it is essential for electricity power supply companies to have accurate short and long-term energy predictions. In this study, a deep-ensembled neural network was used to anticipate hourly power utilization, providing a clear and effective approach for predicting power consumption. The dataset comprises of 13 files, each representing a different region, and ranges from 2004 to 2018, with two columns for the date, time, year and energy expenditure. The data was normalized using minmax scalar, and a deep ensembled (long short-term memory and recurrent neural network) model was used for energy consumption prediction. This proposed model effectively trains long-term dependencies in sequence order and has been assessed using several statistical metrics, including root mean squared error (RMSE), relative root mean squared error (rRMSE), mean absolute bias error (MABE), coefficient of determination (R2), mean bias error (MBE), and mean absolute percentage error (MAPE). Results show that the proposed model performs exceptionally well compared to existing models, indicating its effectiveness in accurately predicting energy consumption.

Analogue and Physical Reservoir Computing Using Water Waves: Applications in Power Engineering and Beyond

“…For example, since physical RC systems can naturally approximate complex dynamical systems, they may provide a unified analogue-computer-like modelling framework for the simulation, control and optimisation of time-critical operations in complex multi-component and multi-scaled electric networks [148][149][150][151]. RC systems in general and physical RC systems in particular have also been efficient in modelling highly nonlinear processes often encountered in energy research [62][63][64][65][66][67][68][69]152] and the predictive control of industrial processes [153]. However, as mentioned previously, RC systems can be readily hybridised with deep learning methods, resulting in a combined approach that benefits from the strengths of deep learning, of ESN and LSM and of the hardware implementation of these algorithms [30].…”

“…We also note the remarkable versatility of RC systems, independently of their implementation as a computer program or hardware physical system. In particular, the standard ESN algorithm and its modifications have been used to predict energy consumption and power generation [62][63][64][65][66][67][68][69], thereby extending more traditional modelling approaches such as vector regression models [70]. Thus, the discussion presented below will be of interest to readers with a background in power engineering and adjacent areas, also giving comprehensive information for those seeking to better understand the field of physical reservoir computing and the opportunities that it can bring.…”