基于Logistic回归模型的电力新能源频率响应方法

Frequency Response Method for Power Renewable Energy Based on Logistic Regression Model

常规的电力新能源频率响应方法以频率波动提取为主,易受到频率二次跌落的影响,导致频率响应效果不佳。因此,设计了一种基于Logistic回归模型的电力新能源频率响应方法。首先提取电力新能源共模频率响应降阶特征,从电力新能源频率数据中提取与频率响应相关的特征,简化回归模型处理的复杂性;然后基于Logistic回归模型建立频率响应模式矩阵,映射新能源发电功率与频率响应之间的复杂关系;最后量化电力新能源频率响应限制功率,根据等效惯性常数、负荷变化等指标,确定新能源发电产生的不平衡功率,从而维系电力系统频率的稳定性。试验结果表明,该方法的频率响应效果较佳,具有实际应用价值。

Traditional frequency response methods for power renewable energy focus on frequency fluctuation extraction,which is often affected by secondary frequency drops,leading to suboptimal response effects.This study proposes a frequency response method based on the Logistic Regression model.By extracting reduced-order characteristics of common-mode frequency response of power renewable energy and extracte the features related to frequency response from renewable energy frequency data and simplifying the regression model's complexity.The we establish a frequency response model matrix using Logistic Regression to map the complex relationship between renewable energy generation power and frequency response.It quantifies the frequency response limits of renewable energy power and determines the imbalance power generated,helping to maintain system frequency stability.Experimental results show that this method offers improved frequency response performance and is applicable in practical scenarios.