Andrej F. Gubina - University of Ljubljana, Faculty of Electrical Engineering, Slovenia
Suad S. Halilcevic - University of Tuzla, Faculty of Electrical Engineering Bosnia and Herzegovina
Ferdinand. Gubina - University of Ljubljana, Faculty of Electrical Engineering, Slovenia

To predict the security level of the power system, Markov chains in conjunction with Monte Carlo simulation are used in the paper. In the new approach, power system security states are classified to discrete ranges. Using a Markov chain for each security range, the temporal development of security level is tracked. Based on the forecasted and recorded data, the proposed simple and computationally fast method offers a useful tool to the system operator to predict the future power system security level, including the probability of a power system blackout. The method has been successfully tested on the Bosnian power system model using actual security level data. The method could be incorporated in the wide-area monitoring system in control centers.

Power system security, Markov chains, Monte Carlo simulation, security prediction