面向容量与输电费用削减的同步峰值预测
Coincident Peak Prediction for Capacity and Transmission Charge Reduction
AI总结：
本文针对电网容量与输电费用分配中的同步峰值（CP）事件预测问题，提出基于场景生成和蒙特卡洛估计的方法，并通过自适应阈值策略回测验证其在电池储能系统负荷削减中的应用价值。
摘要：
满足电网不断增长的需求需要持续的基础设施增强。电网确保向消费者持续可靠供电的能力有两个重要方面：容量，即系统能够处理的最大电量，以及输电，即跨网络输送电力所需的基础设施。这些容量和输电费用随后根据成本 causation 原则分配给最终用户。这些费用根据客户在同步峰值（CP）事件期间的需求计算，CP 事件是系统范围内电力负荷最高的时间区间。我们解决基于不同辖区实际负荷和预测数据预测 CP 事件的问题。特别地，我们根据是否可获得预测数据识别出两种主要用例。我们的方法生成场景，并构建蒙特卡洛估计器，用于预测 CP 日和精确 CP 小时事件。最后，我们回测了具有自适应阈值的策略在预测任务上的性能。该分析使我们能够推导出通过电池储能系统（BESS）解决方案进行负荷削减的实际启示。
英文摘要：
Meeting the ever-growing needs of the power grid requires constant infrastructure enhancement. There are two important aspects for a grid ability to ensure continuous and reliable electricity delivery to consumers: capacity, the maximum amount the system can handle, and transmission, the infrastructure necessary to deliver electricity across the network. These capacity and transmission costs are then allocated to the end-users according to the cost causation principle. These charges are computed based on the customer demand on coincident peak (CP) events, time intervals when the system-wide electric load is highest. We tackle the problem of predicting CP events based on actual load and forecast data on the load of different jurisdictions. In particular, we identify two main use cases depending on the availability of a forecast. Our approach generates scenarios and formulates Monte-Carlo estimators for predicting CP-day and exact CP-hour events. Finally, we backtest the prediction performance of strategies with adaptive threshold for the prediction task. This analysis enables us to derive practical implications for load curtailment through Battery Energy Storage System (BESS) solutions.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
