A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN

Zhu, Yong; Pu, Liangyi; Yang, Di; Kang, Tun; Liang, Chao; Peng, Mingzhi; Zhai, Chao

doi:10.3390/en18215598

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN

by

Yong Zhu

¹,

Liangyi Pu

¹,

Di Yang

¹,

Tun Kang

²,

Chao Liang

¹,

Mingzhi Peng

^3,* and

Chao Zhai

⁴

¹

Chongqing Huizhi Energy Co., Ltd., Chongqing 400000, China

²

SPIC Chongqing Co., Ltd., Chongqing 400000, China

³

School of Future Technology, China University of Geosciences, Wuhan 430074, China

⁴

School of Automation, China University of Geosciences, Wuhan 430074, China

^*

Author to whom correspondence should be addressed.

Energies 2025, 18(21), 5598; https://doi.org/10.3390/en18215598 (registering DOI)

Submission received: 12 September 2025 / Revised: 18 October 2025 / Accepted: 22 October 2025 / Published: 24 October 2025

(This article belongs to the Section F5: Artificial Intelligence and Smart Energy)

Download Versions Notes

Abstract

Virtual Power Plant (VPP) is capable of aggregating and intelligently coordinating diverse distributed energy resources, among which the accuracy of load forecasting is a key factor in ensuring their regulation capability. To address the periodicity and complex nonlinear fluctuations of electricity load data, this study introduces a Cyclic Order Mapping (COM) encoding method, which maps weekly and intraday sequences into continuous ordered variables on the unit circle, thereby effectively preserving load periodic features. On the basis of the COM encoding, a novel forecasting model is proposed by integrating Bidirectional Long Short-Term Memory (BiLSTM) networks, an efficient self-attention mechanism, and the Kolmogorov–Arnold Network (KAN). This model is termed BiLSTM-Att-KAN. Comparative and ablation experiments were conducted to assess the scientific validity and predictive accuracy of the proposed approach. The results confirm its superiority, achieving a Root Mean Square Error (RMSE) of 141.403, a Mean Absolute Error (MAE) of 106.687, and a coefficient of determination (R²) of 0.962. These findings demonstrate the effectiveness of the proposed model in enhancing load forecasting performance for VPP applications.

Keywords: virtual power plants; COM encoding; electrical load forecasting; BiLSTM-Att-KAN

Share and Cite

MDPI and ACS Style

Zhu, Y.; Pu, L.; Yang, D.; Kang, T.; Liang, C.; Peng, M.; Zhai, C. A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN. Energies 2025, 18, 5598. https://doi.org/10.3390/en18215598

AMA Style

Zhu Y, Pu L, Yang D, Kang T, Liang C, Peng M, Zhai C. A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN. Energies. 2025; 18(21):5598. https://doi.org/10.3390/en18215598

Chicago/Turabian Style

Zhu, Yong, Liangyi Pu, Di Yang, Tun Kang, Chao Liang, Mingzhi Peng, and Chao Zhai. 2025. "A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN" Energies 18, no. 21: 5598. https://doi.org/10.3390/en18215598

APA Style

Zhu, Y., Pu, L., Yang, D., Kang, T., Liang, C., Peng, M., & Zhai, C. (2025). A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN. Energies, 18(21), 5598. https://doi.org/10.3390/en18215598

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

A Virtual Power Plant Load Forecasting Approach Using COM Encoding and BiLSTM-Att-KAN

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI