| 引用本文: | 肖白,邢世亨,王茂春,杨森林,苟晓侃.基于改进KDE法和GA-SVM的多风电场聚合后输出功率长期波动特性预测方法[J].电力自动化设备,2022,42(2): |
| XIAO Bai,XING Shiheng,WANG Maochun,YANG Senlin,GOU Xiaokan.Prediction method of output power long-term fluctuation characteristic for multiple wind farms after aggregation based on improved KDE method and GA-SVM[J].Electric Power Automation Equipment,2022,42(2): |
|
| 摘要: |
| 针对规划期内有新增风电装机容量但没有与其对应的实测风电输出功率数据，导致难以准确把握和刻画规划目标年多风电场聚合后输出功率长期波动特性的问题，提出一种利用改进核密度估计(KDE)法和经遗传算法寻优的支持向量机(GA-SVM)预测多风电场聚合后输出功率长期波动特性的方法。对风电功率的长期波动特性进行刻画，分析在多风电场聚合过程中装机容量与风电功率之间的关系；运用改进KDE法生成多风电场聚合过程中不同装机容量下的输出功率概率密度曲线；采用GA-SVM建立多风电场聚合后输出功率概率密度演变模型；根据概率分布与持续功率曲线的对应关系，对预测出的规划目标年的多风电场聚合后的输出功率概率密度曲线进行反演，得到可描述规划目标年输出功率长期波动特性的持续功率曲线。工程实例证明了所提方法的实用性和有效性。 |
| 关键词: 多风电场 风电波动特性 核密度估计 支持向量机 |
| DOI：10.16081/j.epae.202111015 |
| 分类号:TM714;TM614 |
| 基金项目:国家重点研发计划项目(2017YFB0902200) |
|
| Prediction method of output power long-term fluctuation characteristic for multiple wind farms after aggregation based on improved KDE method and GA-SVM |
|
XIAO Bai1, XING Shiheng2, WANG Maochun3, YANG Senlin3, GOU Xiaokan3
|
|
1.School of Electrical Engineering, Northeast Electric Power University, Jilin 132012, China;2.Yanbian Power Supply Company of State Grid Jilin Electric Power Supply Co.,Ltd.,Yanji 133000, China;3.State Grid Qinghai Electric Power Company, Xining 810008, China
|
| Abstract: |
| Aiming at the problem that there exists new-added wind power installed capacity during the planning period but its corresponding measured wind power output data is lacked, which causes the long-term fluctuation characteristic of output power of multiple wind farms after aggregation in the planning target year is difficult to be accurately grasped and described, a prediction method for long-term fluctuation characteristic of output power of multiple wind farms after aggregation is proposed based on improved KDE(Kernel Density Estimation) method and GA-SVM(Support Vector Machine optimized by Genetic Algorithm). The long-term fluctuation characteristic of output power of wind power is described, and the relationship between installed capacity and wind power is analyzed during the aggregation process of multiple wind farms. The improved KDE method is used to generate the probability density curves of output power during the aggregation process of multiple wind farms with different installed capacities. GA-SVM is adopted to establish the probability density varying model of output power after aggregation of multiple wind farms. According to the corresponding relationship between probability distribution and duration power curve, the predicted probability density curve of output power for multiple wind farms after aggregation in the planning target year is inversed so that the duration power curve which can describe the long-term fluctuation characteristic of output power in the planning target year is obtained. Engineering project verifies the practicability and effectiveness of the proposed method. |
| Key words: multiple wind farms wind power fluctuation characteristic kernel density estimation support vector machine |
