| 引用本文: | 刘晟,向明旭,刘硕,于松泰,杨知方.分区匹配气象及功率特征的运行备用需求量化[J].电力自动化设备,2024,44(3):150-157 |
| LIU Sheng,XIANG Mingxu,LIU Shuo,YU Songtai,YANG Zhifang.Operating reserve demand quantification based on interval matching of meteorology and power features[J].Electric Power Automation Equipment,2024,44(3):150-157 |
|
| 摘要: |
| 现有运行备用需求量化方法未考虑气象因素对系统整体预测随机性的影响，难以兼顾电网运行安全性和经济性。为此，提出了分区匹配气象及功率特征的运行备用需求量化方法。根据历史数据建立气象-功率二维区间，采用非参数核密度估计方法拟合不同区间内的净负荷预测误差分布，从而刻画不同气象、功率条件下系统整体预测的随机性；提出基于数据特征相似度的历史数据筛选和数据区间划分策略，以提升预测随机性的刻画准确性；根据运行日隶属的二维区间估计其净负荷预测随机性，据此量化给定置信水平下的系统运行备用需求。以我国某省级电网的实际数据为算例，验证了所提方法的有效性。 |
| 关键词: 运行备用 预测随机性 气象因素 非参数核密度估计 特征相似度 |
| DOI：10.16081/j.epae.202307007 |
| 分类号:TM73 |
| 基金项目:北京电力交易中心有限公司科技项目(新能源承担系统消纳成本的电力市场出清模型及定价机制)(SGDJ0000YJJS2200026) |
|
| Operating reserve demand quantification based on interval matching of meteorology and power features |
|
LIU Sheng1, XIANG Mingxu1, LIU Shuo2, YU Songtai2, YANG Zhifang1
|
|
1.State Key Laboratory of Power Transmission Equipment & System Security and New Technology, Chongqing University, Chongqing 400044, China;2.Beijing Power Trading Center Co.,Ltd.,Beijing 100031, China
|
| Abstract: |
| The current operating reserve demand quantification methods do not consider the influence of meteorological factors on the system’s overall forecast randomness, so it is difficult to take into account the safety and economy of power grid operation. Therefore, an operating reserve demand quantification method based on interval matching of meteorology and power features is proposed. Based on the historical data, the meteorology-power two-dimensional intervals are established, and the distribution of net load forecast error in different intervals is fitted by using the non-parametric kernel density estimation method, so as to describe the system’s overall forecast randomness under different meteorological and power conditions. In order to improve the characterization accuracy of forecast randomness, the historical data selection strategy and data interval division strategy based on data feature similarity are proposed. The randomness of net load prediction of the operating day is estimated according to the two-dimensional interval which the operating day belongs to, based on which, the operating reserve demand of the system is quantified under the given confidence levels. Taking the actual data of a provincial power grid in China as an example, the effectiveness of the proposed method is verified. |
| Key words: operating reserve forecasting uncertainty meteorological factors non-parametric kernel density estimation feature similarity |
