Algorithms for fault tolerance in emerging smart power grids

点击翻译按钮获取中文摘要

Fault-tolerant computing plays an important role in today's world. Fault tolerance prevents system failures and is intended to ensure that the system delivers the required service in spite of faults and errors. In the past two decades, we have seen the evolution of computer networks as a merger of the communication and computing worlds. One example is the emerging smart power grids. Cascading failures occur in the power grid when nodes such as transmission lines or distribution feeders in the physical power system fail and their power load is passed to another local node which can result in overload, resulting in cascading effects. Conventional power grids provide very little protection against cascading failures. As demand for power increases, potential inability of the power grid to support the occasional increased load could lead to more blackouts. The blackout of 2003, for example, affected 55 million people and caused 256 power plants to go offline. Intelligent energy and fault management can identify failure, disconnect certain plants, distribute power load and decision making processes, and resort to islanding and isolation from either incoming or outgoing faults. The goal of this project is to develop better analytical models for power grid topologies and cascading failures which will help us design future topologies to limit power cascading failures.