GOALI: Intelligent Leader Follower Systems for Control of Energy Services

GOALI：用于控制能源服务的智能领导者跟随者系统

• 批准号: 0118080
• 负责人: Ali Keyhani
• 金额: $ 14.5万
• 依托单位: Ohio State University Research Foundation -DO NOT USE
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-06-15 至 2005-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0118080&HistoricalAwards=false
• 关键词: GOALI; Intelligent; Leader; Follower; Systems

0118080KeyhaniThe collapse of the California electric energy market has made it imperative to recognize that market forces motivated by profit cannot guarantee a secure and reliable energy supply because of the inherent characteristics of electric energy production, transmission and distribution networks. No other systems, such as transportation, gas distribution, airline, trucking are subjected to high prices because of an outage due to maintenance of one of its components. Therefore, it is essential that the electricity market be designed such that the market players' profit motives will mesh with system security and reliability. This proposal will undertake the study of this problem and possible solutions using the appropriate market structure.The Ohio State University (OSU), Illinois Institute of Technology University (lIT), ABB Energy Information Systems (ABB) and Perot Systems (PS) will undertake fundamental research in system game theory to develop a new market structure for planned operation of electric energy services. The PIs propose to develop algorithms to create an efficient, reliable, and secure energy market with the ISO as a market leader who is also a market maker armed with incentive functions, and followers who compete to provide energy for stable operation of the system. The approach is based on the Stackelberg market design strategy developed in the 1970's. We will design a Virtual Market Simulator (VMS) to evaluate the proposed market structure. The VMS simulator will be used for evaluation of system operation yearly on monthly basis, monthly on a weekly basis, weekly on a daily basis, daily on an hourly basis, and hourly on a minute basis up to AGC cycle. The PIs will use the VMS test bed to demonstrate that a market based on leader- follower algorithms can attain perfect market prices for the energy and ancillary services markets. To achieve this goal, they will develop a VMS simulator test bed in collaboration with their industrial partners, ABB and PS, using ABB and PS Systems' knowledge of PJM and California systems historical data. In their VMS simulator, they will construct neural network based predictors to forecast energy demands and time-varying cost of the system load centers (monthly, weekly, daily and hourly). They will formulate an incentive function for each type of generation that the followers must satisfy in their bid prices. The incentive functions will be determined by the leader, taking into account the rational profit motives of the players as well as the system security and reliability. In such market environments, all market players will be followers. In the proposed market structure, the leader will use the knowledge of the systems that he or she is controlling and will estimate the production cost of each type of generation. The leader will thus have the ability to formulate leader-follower optimization strategies and design an appropriate incentive function for each type of generation system. At the same time the market leader can formulate a cost function for the system operation based on the mandated security and reliability by state power boards. In this market structure, the state power board will determine the overall energy policy and set up an auction market for market players to compete in the market under leader-follower market structure. In our formulation, the market leader position is assigned to ISO's. The significance of the proposed work is the creation of a new market structure that will make it feasible for market forces to participate in the development of electric energy systems and to compete in the market while ensuring the reliability and security of systems. The proposed market structure will allow the state government to formulate energy policy and, through proposed incentive functions, to encourage the citizen to participate in its implementation. The proposed market structure will also help market forces develop energy sources, since it will remove the uncertainties of planning plants where they may face bottlenecks in transmissions. Energy users will also benefit by reducing the cost of energy through participation in a load response program and being assured of reliable and secure service.

加州电能市场的崩溃使得必须认识到，由于电能生产、传输和分配网络的固有特性，受利润驱动的市场力量不能保证安全可靠的能源供应。没有其他系统，如运输，天然气配送，航空公司，卡车运输，因为其组件之一的维护而中断而受到高价。因此，电力市场的设计必须使市场参与者的利润动机与系统的安全性和可靠性相结合。该提案将对这一问题进行研究，并采用适当的市场结构提出可能的解决方案。俄亥俄州州立大学（OSU）、伊利诺伊理工大学（lIT）、ABB能源信息系统公司（ABB）和佩罗系统公司（PS）将进行系统博弈论的基础研究，为电力能源服务的计划运营开发新的市场结构。PI建议开发算法，以创建一个高效，可靠和安全的能源市场，ISO作为市场领导者，也是一个拥有激励功能的做市商，以及竞争为系统稳定运行提供能源的追随者。该方法是基于20世纪70年代开发的Stackelberg市场设计策略。我们将设计一个虚拟市场模拟器（VMS）来评估拟议的市场结构。VMS模拟器将用于每年每月、每月每周、每周每天、每天每小时和每分钟每小时评估系统运行，直至AGC周期。PI将使用VMS测试床来证明基于领导者-追随者算法的市场可以为能源和辅助服务市场实现完美的市场价格。为了实现这一目标，他们将与工业合作伙伴ABB和PS合作，利用ABB和PS Systems对PJM和加州系统历史数据的了解，开发一个VMS模拟器测试台。在VMS模拟器中，他们将构建基于神经网络的预测器，以预测系统负荷中心的能源需求和随时间变化的成本（每月、每周、每日和每小时）。他们将为每种发电类型制定一个激励函数，追随者必须在他们的投标价格中满足这个激励函数。激励函数将由领导者确定，同时考虑到参与者的合理利润动机以及系统的安全性和可靠性。在这样的市场环境下，所有的市场参与者都将成为追随者。在拟议的市场结构中，领导者将利用他或她所控制的系统的知识，并估计每种发电类型的生产成本。因此，领导者将有能力制定领导者-跟随者优化策略，并为每种类型的发电系统设计适当的激励函数。同时，市场领导者可以根据国家电力局规定的安全性和可靠性制定系统运行的成本函数。在这种市场结构下，国家电力局将决定整体的能源政策，并建立一个拍卖市场，供市场参与者在领导者-追随者的市场结构下进行竞争。在我们的制定中，市场领导者的位置被分配给ISO。拟议工作的意义在于建立一个新的市场结构，使市场力量能够参与电力能源系统的发展，并在市场上竞争，同时确保系统的可靠性和安全性。拟议的市场结构将允许州政府制定能源政策，并通过拟议的激励功能，鼓励公民参与其执行。拟议的市场结构还将有助于市场力量开发能源，因为它将消除规划发电厂时可能面临传输瓶颈的不确定性。能源用户也将受益于通过参与负荷响应计划降低能源成本，并确保可靠和安全的服务。