Online stability assessment for isolated microgrid via LASSO based neural network algorithm

IF 7.1 Q1 ENERGY & FUELS

Energy Conversion and Management-X Pub Date : 2025-01-01 DOI:10.1016/j.ecmx.2024.100849

Ahmed Lasheen , Hatem F. Sindi , Hatem H. Zeineldin , Mohammed Y. Morgan

引用次数: 0

Abstract

Online prediction of the dominant modes is very important for microgrid operation. The dominant modes determine microgrid stability and the active and reactive power oscillations. Therefore, online prediction of these modes is essential to check the microgrid stability periodically. Consequently, this paper introduces an artificial intelligent algorithm to identify the dominant modes of the microgrid. This algorithm combines a cascaded feedforward neural network with the least absolute shrinkage and select operator (LASSO). The LASSO algorithm is used to extract the most important data that affects the dominant modes. On the other hand, the cascaded feedforward neural network is trained using LASSO data to identify the microgrid dominant modes. The proposed algorithm is tested using a 6-bus AC microgrid. The results show that the proposed algorithm significantly determines the dominant modes of the microgrid by using a minimum set of data determined by LASSO.

查看原文本刊更多论文

在线预测主导模式对于微电网的运行非常重要。主导模式决定了微电网的稳定性以及有功和无功功率的振荡。因此，这些模式的在线预测对于定期检查微电网的稳定性至关重要。因此，本文引入了一种人工智能算法来识别微电网的主导模式。该算法将级联前馈神经网络与最小绝对收缩和选择算子（LASSO）相结合。LASSO 算法用于提取影响主导模式的最重要数据。另一方面，利用 LASSO 数据训练级联前馈神经网络，以识别微电网主导模式。使用 6 总线交流微电网对所提出的算法进行了测试。结果表明，通过使用 LASSO 确定的最小数据集，所提出的算法能显著确定微电网的主导模式。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Energy Conversion and Management-X Multiple-

CiteScore

8.80

自引率

3.20%

发文量

180

审稿时长

58 days

期刊介绍： Energy Conversion and Management: X is the open access extension of the reputable journal Energy Conversion and Management, serving as a platform for interdisciplinary research on a wide array of critical energy subjects. The journal is dedicated to publishing original contributions and in-depth technical review articles that present groundbreaking research on topics spanning energy generation, utilization, conversion, storage, transmission, conservation, management, and sustainability. The scope of Energy Conversion and Management: X encompasses various forms of energy, including mechanical, thermal, nuclear, chemical, electromagnetic, magnetic, and electric energy. It addresses all known energy resources, highlighting both conventional sources like fossil fuels and nuclear power, as well as renewable resources such as solar, biomass, hydro, wind, geothermal, and ocean energy.