Hybrid energy storage compensation strategy for photovoltaic systems based on predictive error analysis

Read full text Summarise with VIRAT
Authors: Daosheng Liu, Ruimin Yang, Zijian Zhang, Huanjing Qiu
Cited by: 0
References: 2

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

To address the power prediction errors caused by random fluctuations in photovoltaic output, a hybrid energy storage compensation strategy integrating signal optimization decomposition and frequency-domain analysis is proposed. Fuzzy C-means clustering and Bidirectional Gated Recurrent Unit are employed to extract error data under different feature types, while the Sparrow Search Algorithm is utilized to optimize the parameters of the Variational Mode Decomposition algorithm, enabling accurate multi-scale decomposition of prediction errors. Subsequently, the decomposed signals are analyzed in the frequency domain using the Hilbert marginal spectrum to determine the frequency division points of various error components. Based on these results, frequency-domain coordinated compensation between energy-type and power-type storage units is achieved, ensuring complementary and collaborative operation of the hybrid energy storage system. Experimental results demonstrate that the proposed frequency-domain-based hybrid compensation strategy significantly enhances the compensation efficiency of the energy storage system. Moreover, comparison through the comprehensive daily degradation index indicates that the proposed method effectively reduces battery degradation and improves the overall economic performance of the system.

Fields of Study & Subjects

Power Systems and Renewable Energy Photovoltaic System Optimization Techniques Smart Grid Energy Management

Publication details

Journal / source
Journal of Physics Conference Series
Volume
3198
Issue
1
Pages
012043–012043
Publication date
1 March 2026
Language
EN