Probability models for estimating the probabilities of cascading outages in high-voltage transmission network

This paper discusses a number of probability models for multiple transmission line outages in power systems, including generalized Poisson model, negative binomial model, and exponentially accelerated model. These models are applied to the multiple transmission outage data for a 20-year period for North America. The probabilities of the propagation of transmission cascading outage are calculated. These probability magnitudes can serve as indexes for long-term planning and can also be used in short-term operational defense to such events. Results from our research show that all three models apparently explain the occurrence probability of higher order outages very well. However, the exponentially accelerated model fits the observed data and predicts the acceleration trends best. Strict chi-squared fitness tests were done to compare the fitness among these three models, and the test results are consistent with what we observe.