{"title":"Self-paced learning long short-term memory based on intelligent optimization for robust wind power prediction","authors":"Shun Yang, Xiaofei Deng, Dongran Song","doi":"10.1049/cth2.12644","DOIUrl":null,"url":null,"abstract":"<p>Given the unpredictable and intermittent nature of wind energy, precise forecasting of wind power is crucial for ensuring the safe and stable operation of power systems. To reduce the influence of noise data on the robustness of wind power prediction, a wind power prediction method is proposed that leverages an enhanced multi-objective sand cat swarm algorithm (MO-SCSO) and a self-paced long short-term memory network (spLSTM). First, the actual wind power data is processed into time series as input and output. Then, the progressive advantage of self-paced learning is used to effectively solve the instability caused by noisy data during long short-term memory network (LSTM) training. Following this, the improved MO-SCSO is employed to iteratively optimize the hyperparameters of spLSTM. Ultimately, a combined MO-SCSO-spLSTM model is constructed for wind power prediction. This model is validated with the data of onshore wind farms in Austria and offshore wind farms in Denmark. The experimental results show that compared with the traditional LSTM prediction method, the proposed method has better prediction accuracy and robustness. Specifically, in the onshore and offshore wind power prediction experiments, the proposed method reduces the minimum MAE by 5.44% and 4.96%, respectively, and reduces the MAE range by 4.45% and 17.21%, respectively, which could be conducive to the safe and stable operation of power system.</p>","PeriodicalId":50382,"journal":{"name":"IET Control Theory and Applications","volume":"18 17","pages":"2239-2255"},"PeriodicalIF":2.2000,"publicationDate":"2024-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/cth2.12644","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IET Control Theory and Applications","FirstCategoryId":"94","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1049/cth2.12644","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"AUTOMATION & CONTROL SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
Given the unpredictable and intermittent nature of wind energy, precise forecasting of wind power is crucial for ensuring the safe and stable operation of power systems. To reduce the influence of noise data on the robustness of wind power prediction, a wind power prediction method is proposed that leverages an enhanced multi-objective sand cat swarm algorithm (MO-SCSO) and a self-paced long short-term memory network (spLSTM). First, the actual wind power data is processed into time series as input and output. Then, the progressive advantage of self-paced learning is used to effectively solve the instability caused by noisy data during long short-term memory network (LSTM) training. Following this, the improved MO-SCSO is employed to iteratively optimize the hyperparameters of spLSTM. Ultimately, a combined MO-SCSO-spLSTM model is constructed for wind power prediction. This model is validated with the data of onshore wind farms in Austria and offshore wind farms in Denmark. The experimental results show that compared with the traditional LSTM prediction method, the proposed method has better prediction accuracy and robustness. Specifically, in the onshore and offshore wind power prediction experiments, the proposed method reduces the minimum MAE by 5.44% and 4.96%, respectively, and reduces the MAE range by 4.45% and 17.21%, respectively, which could be conducive to the safe and stable operation of power system.
期刊介绍:
IET Control Theory & Applications is devoted to control systems in the broadest sense, covering new theoretical results and the applications of new and established control methods. Among the topics of interest are system modelling, identification and simulation, the analysis and design of control systems (including computer-aided design), and practical implementation. The scope encompasses technological, economic, physiological (biomedical) and other systems, including man-machine interfaces.
Most of the papers published deal with original work from industrial and government laboratories and universities, but subject reviews and tutorial expositions of current methods are welcomed. Correspondence discussing published papers is also welcomed.
Applications papers need not necessarily involve new theory. Papers which describe new realisations of established methods, or control techniques applied in a novel situation, or practical studies which compare various designs, would be of interest. Of particular value are theoretical papers which discuss the applicability of new work or applications which engender new theoretical applications.