EAGER: Adaptive Dynamic Programming for Residential Energy System Control and Management

点击翻译按钮获取中文摘要

This crossdisciplinary proposal aims to apply new intelligent optimization methods to the challenge of intelligent price-responsive management of residential energy use, with an emphasis on home battery use connected to the grid. There has been considerable discussion of the importance of distributed energy storage, including batteries in the home, as a way to create more price-responsive demand and as a way to assimilate more wind power more effectively into the grid. However, to make the best possible use of such batteries for the benefit of the grid, with time of day pricing signals, is a difficult multistage stochastic optimization problem ? also known as a dynamic programming problem. Interest in approximate dynamic programming (ADP) has grown in the power sector, and a few applications have appeared for generators and grid management, but this would be the first application to price-responsive residential demand of any kind. Broad Impacts: This project would be in helping the PI ? a major contributor in ADP, with a strong background in control theory, and strong involvement in IEEE outreach activities and partnerships with China ? make a transition to the intelligent grid areas, where a new type of crossdisciplinary background is badly needed. The need for more storage in the grid is a critical obstacle to greater use and benefit from wind. At present, in the PJM system, a megawatt of wind power requires 0.87 megawatts of new fossil fuel generation to back it up, imposing a huge barrier to its economics and to its use; more storage could eliminate the need for backup. Many are hoping that storage in plug-in hybrid cars can solve this problem, but economically profitable batteries in the home could be deployed more quickly. The necessary 5-minute price signals are available in most leading power regions in the US, and MidWest ISO (where this university is located) has made a special commitment to the kind of price-based management which would make it feasible to deploy this kind of solution.