基于样本数据迁移学习的贫资料地区小水电超短期出力建模及发电预测.pdfVIP

第 43 卷 第 7 期 中 国 电 机 工 程 学 报 Vol.43 No.7 Apr. 5, 2023 2652 2023 年 4 月 5 日 Proceedings of the CSEE ©2023 Chin.Soc.for Elec.Eng. DOI ：10.13334/j.0258-8013.pcsee.212895 文章编号：0258-8013 (2023) 07-2652-14 中图分类号：TM 71 文献标识码：A 基于样本数据迁移学习的贫资料地区小水电 超短期出力建模及发电预测 * 魏泽涛，刘友波，沈晓东 ，刘俊勇 (四川大学电气工程学院，四川省 成都市 610065) Ultra-short-term Power Generation Modeling and Prediction for Small Hydropower in Data-scarce Areas Based on Sample Data Transfer Learning * WEI Zetao, LIU Youbo, SHEN Xiaodong , LIU Junyong (School of electrical engineering, Sichuan University, Chengdu 610065, Sichuan Province, China) ABSTRACT: Modeling the generation capacity of small sample data of data-scarce small hydropower are verified with hydropower in rich areas is an important measure to ensure the an example in a region of Sichuan. Compared with traditional stable and economic operation of the power grid and to realize deep learning, this method can reduce the root mean square the coordination of large and small hydropower. The error of prediction by 16.54% on average, which has practical information collection of small hydropower in remote areas is engineering value. difficult, and modeling-related valid data is scarce, which KEY WORDS: transfer learning; time series data makes it difficult to draw on existing methods for large and segmentation; parameter transfer; small hydropower generation medium-si