AI 中文总结
研究针对随机能源系统优化模型中忽视不确定输入参数统计依赖性的问题，提出基于copula的工作流程来识别、表征和建模相关结构，以奥地利数据为例验证该方法有效，还指出迈向完整框架的后续关键步骤。
AI 中文摘要
在随机能源系统优化模型中，不确定输入参数之间的统计依赖性常被忽视，尽管这会严重影响结果。为填补这一空白，我们正在开发一个用于表征、建模和基准测试统计依赖性的综合框架。在这项工作中，我们提出了一种基于copula的工作流程，以识别、表征和建模输入参数之间的线性和单调相关结构，这是迈向该框架的第一步。我们使用2019年至2025年奥地利的太阳能发电、日前电价和电力需求数据展示了我们的工作流程。结果表明这些变量之间存在显著的线性和单调依赖性，且基于copula的方法能很好地捕捉这种依赖性。最后，基于这个可扩展的基础，我们强调了迈向完整框架的下一步关键杠杆，包括时间依赖性和高维依赖性建模。
英文摘要
Statistical dependence among uncertain input parameters in stochastic energy system optimization models is often ignored, even though this can substantially bias outcomes. To address this gap, we are developing a comprehensive framework for characterizing, modelling, and benchmarking statistical dependence. In this work, we present a copula-based workflow to identify, characterize, and model linear and monotonic correlation structures between input parameters, representing the first development step towards this framework. We demonstrate our workflow using solar generation, day-ahead electricity prices, and electricity demand data in Austria between 2019 and 2025. Our results show substantial linear and monotonic dependence between these variables, and that this dependence is well captured by the copula-based approach. Finally, building on this scalable foundation, we highlight key levers for next steps towards the full framework, including time-dependent and higher-dimensional dependence modelling.
CommentsEuropean Energy Market (EEM) 2026 conference
