Towards an Understanding of Deregulated Electricity Markets through Time Series Analysis

本研究的目的是了解在新的管理法规下电力市场的行为与大型电力系统的工程方面之间的相互作用。在允许金融机制自由地指导电力系统的日常运作方面存在许多实际困难。电力商品必须在严格的物理规律和极高的可靠性要求下实时地产生、分配和使用。这一过程比传统商品（如石油或农产品）的分销要严格和复杂得多。此外，电力系统对所有经济活动领域的重要性表明，对任何可能从根本上改变系统运行的变化都要谨慎。当然，评估解除管制市场的有效性对于维持可靠的电力系统性能非常重要。传统的工程学和经济学观点都不能单独为交易和其他金融工具的行为提供完整的解释。与提议的工作直接相关的是价格变动的可预测模式的存在，这可能表明市场或电力系统的问题。这些问题包括价格波动、系统可靠性差、过度的市场力量导致价格高企以及市场效率低下。本研究将发展技术来分析解除管制的电力市场，以确定是否存在代表电力系统正常运行的关注条件的市场行为。特别地，本研究的第一项任务是应用特定领域的知识和定性数据分析来研究相关的电力系统数据，以便在代表电力和辅助服务价格的时间序列中识别有趣的现象。作为第二项任务，选定的时间序列将被仔细预处理，并使用我们在初步研究中开发的方法进行深入分析。我们将重点关注将已确定的定价机制与电力系统的运行状况联系起来。随后，第三个任务包括对第二个任务中研究的时间序列对应的电力市场进行深入的绩效分析。这项最终任务的目标是解决本研究中的基本问题：电力市场中是否存在可预测的定价行为，这些行为表明市场问题，并可能扭曲电力系统的经济和可靠运行。长期目标不是寻求更高的回报，而是制定更有效率和更有效的市场法规。这个项目是一个跨学科的努力，结合了我们之前在以下方面的研究成果：(a)有效的规划和操作工具，将最新的计算方法应用于电力系统工程问题；(b)通过数据分析、先验知识和从实际复杂领域的例子中学习来精确的知识发现系统。

The objective of this study is to gain understanding of the interactions between the behavior of electricity markets under new management regulations and the engineering aspects of large-scale electric power systems. There are numerous practical difficulties in allowing financial mechanisms the freedom to directing day-today operations of the power system. The electric commodity must be generated, distributed and consumed in real-time under strict physical laws and extremely high reliability requirements. This process is far more exacting and complex then the distribution of traditional commodities, such as, oil or agricultural products. Further, the importance of the electric power system to all areas of economic activity suggests caution for any changes that may radically alter system operation. Certainly, assessing the effectiveness of the deregulated markets is of great importance for maintaining reliable power system performance. Neither traditional engineering nor the economics viewpoint can alone provide complete explanations for behavior in trades and other financial instruments. Of direct relevance to the proposed work is the existence of predictable patterns of price movements that may indicate either market or power system problems. These problems include price volatility, poor system reliability, excessive market power leading to high prices, and market inefficiencies. This study will develop techniques to analyze the deregulated electricity market in order to determine if there exists market behavior that represents conditions of concern for proper operation of the power system. In particular, the first task of the proposed study consists of studying relevant power system data applying domain specific knowledge and qualitative data analysis in order to identify interesting phenomena in the time series that represents electricity and ancillary service prices. As a second task, the selected time series will be carefully pre-processed, and subject to an in-depth analysis using methodology developed in our preliminary studies. We will focus on relating identified pricing regimes to power system operating conditions. Subsequently, the third task consists of an in-depth performance analysis for electricity markets corresponding to the time series studied in the second task. The objective of this fanal task is to address the fundamental question in this research: does predictable pricing behavior exist in the electricity markets that indicate market problems and which might distort economic and reliable operation of the power system. The long-term objective is not to seek higher returns but to develop more efficient and effective market regulations.This proposed project is a cross-disciplinary effort that combines our prior research efforts in the development of: (a) effective planning and operation tools that apply the latest computational methods to power system engineering problems; and (b) accurate knowledge discovery systems through data analysis, prior knowledge, and learning from examples in practical complex domains.