Skip to main content
QUICK REVIEW

[论文解读] Multivariate Probabilistic Forecasting of Intraday Electricity Prices using Normalizing Flows

Eike Cramer, Dirk Witthaut|arXiv (Cornell University)|May 27, 2022
Energy Load and Power Forecasting被引用 5
一句话总结

本文提出了一种基于归一化流的多变量概率预测模型,用于日内电力价格预测,以捕捉相对于日前价格的15分钟价格差的联合分布。该模型在趋势准确性和预测区间宽度方面优于高斯Copula和回归基线模型,且XAI分析表明,近期价格历史和日前价格增量是影响最大的特征。

ABSTRACT

Electricity is traded on various markets with different time horizons and regulations. Short-term intraday trading becomes increasingly important due to the higher penetration of renewables. In Germany, the intraday electricity price typically fluctuates around the day-ahead price of the European Power EXchange (EPEX) spot markets in a distinct hourly pattern. This work proposes a probabilistic modeling approach that models the intraday price difference to the day-ahead contracts. The model captures the emerging hourly pattern by considering the four 15 min intervals in each day-ahead price interval as a four-dimensional joint probability distribution. The resulting nontrivial, multivariate price difference distribution is learned using a normalizing flow, i.e., a deep generative model that combines conditional multivariate density estimation and probabilistic regression. Furthermore, this work discusses the influence of different external impact factors based on literature insights and impact analysis using explainable artificial intelligence (XAI). The normalizing flow is compared to an informed selection of historical data and probabilistic forecasts using a Gaussian copula and a Gaussian regression model. Among the different models, the normalizing flow identifies the trends with the highest accuracy and has the narrowest prediction intervals. Both the XAI analysis and the empirical experiments highlight that the immediate history of the price difference realization and the increments of the day-ahead price have the most substantial impact on the price difference.

研究动机与目标

  • 应对由于可再生能源渗透率上升和市场波动性增加所带来的日内电力价格预测挑战。
  • 开发一种多变量概率模型,以捕捉相对于日前价格的四个15分钟日内价格区间的联合分布。
  • 通过利用带条件输入特征的深度生成建模,提升预测准确性和不确定性量化能力。
  • 利用可解释人工智能(XAI)研究外部因素对价格差的影响,识别关键驱动因素。
  • 证明归一化流在短期预测中优于传统概率模型(如高斯Copula和多变量回归)的优越性。

提出的方法

  • 将日内价格差建模为相对于每个交易时段日前价格的四维联合概率分布。
  • 使用归一化流——一种基于可逆神经网络的深度生成模型——来学习在给定外部输入条件下的价格差的条件多变量密度。
  • 将归一化流基于时间段、近期价格差和日前价格增量进行条件化,以实现上下文感知的概率预测。
  • 应用可逆变换将复杂且非高斯的价格差分布映射到标准正态潜空间,从而实现高效的密度估计。
  • 集成可解释人工智能(SHAP值)以分析特征重要性,并解释不同输入条件下模型的行为。
  • 使用标准的概率预测评估指标,将归一化流模型与历史数据选择基线、高斯Copula模型和多变量高斯回归模型进行比较。
Figure 1 : Average daily profiles of day-ahead and ID 3 price trends of 2018 and 2019. EPEX spot price data from Fraunhofer-Institut für Solare Energiesysteme ISE, ( 2022 ) .
Figure 1 : Average daily profiles of day-ahead and ID 3 price trends of 2018 and 2019. EPEX spot price data from Fraunhofer-Institut für Solare Energiesysteme ISE, ( 2022 ) .

实验结果

研究问题

  • RQ1归一化流模型能否有效捕捉相对于日前价格的复杂、非线性和多变量的日内电力价格差结构?
  • RQ2包含外部特征(如近期价格历史和日前价格增量)是否能提升概率预测性能?
  • RQ3不同输入特征对模型预测准确性和不确定性量化能力的相对贡献如何?
  • RQ4在预测锐度和校准性方面,归一化流模型相较于经典概率模型(高斯Copula、多变量回归)表现如何?
  • RQ5短期预测提前时间(如1小时)对模型性能的影响程度如何,相较于历史基线模型?

主要发现

  • 归一化流模型在所有评估模型中实现了最高的趋势准确性,优于历史选择基线以及高斯Copula和回归模型。
  • 该模型生成了最窄的预测区间,表明其不确定性量化能力更强,预测更加锐利。
  • XAI分析识别出价格差的即时历史实现和日前价格的增量是两个最具影响力的输入特征。
  • 当预测提前时间超过一小时后,该模型的性能优势减弱,表明短期动态对准确预测最为关键。
  • 高斯Copula和多变量高斯回归模型未能超越历史选择基线,表明其在建模复杂多变量依赖关系方面存在局限性。
  • 归一化流实现了端到端的、非参数化的联合价格差分布学习,无需对数据的参数形式做出先验假设。
Figure 2 : Box plots (Waskom,, 2021 ) of the energy score (Gneiting and Raftery,, 2007 ; Pinson and Girard,, 2012 ) and variogram score (Scheuerer and Hamill,, 2015 ) distributions of historical data selected using univariate and multivariate approaches. Price data from January 2018 to June 2019 (Fr
Figure 2 : Box plots (Waskom,, 2021 ) of the energy score (Gneiting and Raftery,, 2007 ; Pinson and Girard,, 2012 ) and variogram score (Scheuerer and Hamill,, 2015 ) distributions of historical data selected using univariate and multivariate approaches. Price data from January 2018 to June 2019 (Fr

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。