Steven Rovnyak

Assistant Professor

Key Words:

Electric Power system transient stability

Research description:

Research Areas/Interests: My research has developed new ways to exploit computational resources for real-time transient stability prediction and discrete-event control in electric power systems. Power system protection and remedial action control schemes have traditionally relied on relatively small numbers of simulations which are transformed into decision rules by engineers. My own research uses pattern recognition techniques such as artificial neural networks and decision trees to automatically derive switching rules based on large numbers of simulations. This approach opens the possibility of producing decision rules specific to the current operating condition and continually updating the decision rules over time. It is straightforward to run the separate time-domain simulations in parallel. The computational methods can in fact be applied to fully detailed models of an electric power system, which is one of their main advantages.

Return to University Research Capability