Page 222 - 2024-Vol20-Issue2

218 | Gochhait, Sharma & Bachute

residual block (DCRB), a bidirectional long short-term mem- ACKNOWLEDGMENT
ory layer (Bi-LSTM), and ensemble thinking, outperformed
CNN- based models in accurately forecasting electricity load. This work was supported by the Department of Scientific and
The model’s feature extraction capabilities extend beyond load Industrial Research (DSIR), Government of India under Grant
forecasting andcan contribute to optimal energy allocation, A2KS; Grant number A2KS -11011/7/2022-IRD (SC)- DSIR.
demand-side management, and smart grid operation. Accurate
load forecasting enhances resource optimization, blackout risk CONFLICT OF INTEREST
mitigation, and power supply reliability. Moreover, the model
holds potential for applications in other domains such as stock The authors have no conflict of relevant interest to this article
price forecasting, weather prediction, and traffic flow predic- can be used.
tion. Nevertheless, this research has limitations, including
reliance on a single dataset that may not represent all power REFERENCES
systems and the model’s requirement for extensive training
and computational complexity, limiting its realtime applica- [1] M. Kharrich, L. Abualigah, S. Kamel, H. AbdEl-Sattar,
tion. Future work involves exploring alternative deeplearning and M. Tostado-Ve´liz, “An improved arithmetic opti-
methods and optimizing the model’s parameters to enhance mization algorithm for design of a microgrid with energy
accuracy. Further analysis of factors influencing electricity storage system: Case study of el kharga oasis, egypt,”
load will deepen the understanding of time series data, while Journal of Energy Storage, vol. 51, p. 104343, 2022.
applyingthe model to different power systems and datasets
will assess its generalizability and robustness. In conclusion, [2] L. Wong, V. Ramachandaramurthy, P. Taylor,
this comparative analysis highlights the superiority of the pro- J. Ekanayake, S. Walker, and S. Padmanaban, “Review
posed hybrid deep learning model over existing CNN-based on the optimal placement, sizing and control of an
models for long-term load forecasting in Orissa and Delhi. energy storage system in the distribution network,”
The findings contribute significantly to power systemmanage- Journal of Energy Storage, vol. 21, pp. 489–504, 2019.
ment, and further research can focus on improving accuracy
and expanding the model’s applications across various do- [3] B. Khaki, “Joint sizing and placement of battery energy
mains. storage systems and wind turbines considering reactive
power support of the system,” Journal of Energy Storage,
ACRONYMS AND ABBREVIATIONS vol. 35, p. 102264, 2021.

Long-Term Load Forecasting LTLF [4] S. Mahmoudi, A. Maleki, and D. Ochbelagh, “A novel
Densely Connected Residual Block DCRB method based on fuzzy logic to evaluate the storage and
Root Mean Square Error RMSE backup systems in determining the optimal size of a
Northern Region Load Dispatch Center NRLDC hybrid renewable energy system,” Journal of Energy
Odisha Power Transmission Corporation Lim- OPTCL Storage, vol. 49, p. 104015, 2022.
ited
Megawatts MW [5] M. Nazir, A. Abdalla, H. Zhao, Z. Chu, H. Nazir,
One-Dimensional Convolutional Neural Net- 1D CNN M. Bhutta, and P. Sanjeevikumar, “Optimized economic
work operation of energy storage integration using improved
Bidirectional LSTM BI LSTM gravitational search algorithm and dual stage optimiza-
Short-Term Load Forecasting STLF tion,” Journal of Energy Storage, vol. 50, p. 104591,
Autoregressive Integrated Moving Average ARIMA 2022.
Temperature at 2 Meters (C) T2M
Relative Humidity at 2 Meters (%) RH2M [6] A. Ali, M. Elmarghany, M. Abdelsalam, M. Sabry, and
Specific Humidity at 2 Meters (g/kg) QV2M A. Hamed, “Closed-loop home energy management sys-
Wind Speed at 10 Meters (m/s) WS10M tem with renewable energy sources in a smart grid:
Wet Bulb Temperature at 2 Meters (C) T2MWET A comprehensive review,” Journal of Energy Storage,
Dew/Frost Point at 2 Meters (C) T2MDEW vol. 50, p. 104609, 2022.
Artificial Intelligence AI
Convolutional Neural Networks CNNs [7] A. Maleki, “Design and optimization of autonomous
Recurrent Neural Networks RNNs solar-wind-reverse osmosis desalination systems cou-
Rectified Linear Unit ReLU pling battery and hydrogen energy storage by an im-
proved bee algorithm,” Desalination, vol. 435, pp. 221–
234, 2018.