| 引用本文: | 史林军,戴滔,劳文洁,吴峰,林克曼,李杨,朱玲,黄锡芳.基于改进KNN算法的新能源发电单元运行状态识别[J].电力自动化设备,2024,44(5):65-72. |
| SHI Linjun,DAI Tao,LAO Wenjie,WU Feng,LIN Keman,LI Yang,ZHU Ling,HUANG Xifang.Operating state recognition of new energy power generation unit based on improved KNN algorithm[J].Electric Power Automation Equipment,2024,44(5):65-72. |
|
| 摘要: |
| 目前识别发电单元运行状态的研究较少，数据来源以数据采集与监控系统为主，采集速度较慢。为此，提出了一种基于发电单元机端电气量数据并融合改进k近邻(KNN)算法的新能源发电单元状态识别方法，直接采集机端电气量数据用于快速判断发电单元状态。提出KNN算法的改进策略，克服了传统KNN算法准确度低、识别速度慢的缺点。利用电力系统分析综合程序获取用于状态识别的发电单元机端电气量数据，利用改进策略对数据进行预处理，并对比传统KNN算法、逐条使用改进策略的KNN算法对新能源发电单元状态识别的耗时与准确度。结果表明所提算法较传统算法的识别准确度和速度明显提升，能满足稳定控制过程中对新能源发电单元的状态感知需求。 |
| 关键词: 状态识别 改进KNN算法 新能源发电单元 特征提取 特征加权 |
| DOI：10.16081/j.epae.202311009 |
| 分类号: |
| 基金项目:国家电网公司科技项目(5100-202140478A-0-5-ZN) |
|
| Operating state recognition of new energy power generation unit based on improved KNN algorithm |
|
SHI Linjun1, DAI Tao1, LAO Wenjie1, WU Feng1, LIN Keman1, LI Yang1, ZHU Ling2, HUANG Xifang2
|
|
1.College of Energy and Electrical Engineering, Hohai University, Nanjing 210098, China;2.NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 211106, China
|
| Abstract: |
| At present, there are few researches on the operating state recognition of power generation units, and the data source is mainly supervisory control and data acquisition system, which has a slow acquisition speed. Therefore, an operating state recognition method of new energy power generation units is proposed based on the generator terminal electrical data and the improved k nearest neighbor(KNN) algorithm, which collects the generator terminal electrical data directly to judge the state of the power generation units quickly. An improved strategy of KNN algorithm is proposed to overcome the shortcomings of traditional KNN algorithm, such as low accuracy and slow recognition speed. The power system analysis software package is used to obtain the generator terminal electrical data of the power generation units for state reco-gnition, and the improved strategy is used to preprocess the data. The time-consuming and accuracy of traditional KNN algorithm and improved KNN algorithm for state recognition of new energy power generation units is compared. The results show that the proposed algorithm can improve the recognition accuracy and speed significantly compared with the traditional algorithm, and can meet the state perception requirement of new energy power generation units in the stability control process. |
| Key words: state recognition improved KNN algorithm new energy power generation unit feature extraction feature weighting |
