Title :
In this issue - Technically
Author_Institution :
Indiana Univ. Univ. Indianapolis, Indianapolis
Abstract :
This is a valuable paper that develops pattern recognition methods for real-time transient stability prediction and control. The authors of the present paper have independently developed the technique of labeling all the training data points from unstable simulations as unstable. Labeling a data point as unstable refers to setting the "target" or desired output for that point to unstable.
Keywords :
decision trees; pattern recognition; power system transient stability; controlled islanding; decision tree; pattern recognition methods; real time transient stability prediction; training data points; unstable simulations; Computational modeling; Event detection; Labeling; Pattern recognition; Power system simulation; Predictive models; Protection; Stability; Time measurement; Training data;
Journal_Title :
Power Systems, IEEE Transactions on
DOI :
10.1109/TPWRS.2007.908377
