Classification of input and output variables for a Bayesian model to analyze animal-related outages in overhead distribution systems

Animals, such as squirrels, cause significant outages in overhead distribution systems. Models that would accurately estimate outages caused by animals would be very useful for utilities for year-end analysis of reliability performance of the distribution system. Large increase in outages caused by animals would require the utility to do further evaluation and take remedial actions. A two-layer Bayesian network model with Month-Type, Level of Fair Weather Days in the week, and Outage Level in the previous week as input and Outage Level in the week is presented in this paper for estimation of weekly animal-related outages. Results of different approaches for classification of inputs and output are presented, which are then compared to select the best classification of input and output variables for the model.

[1]  A. Pahwa,et al.  A Probabilistic Approach for Animal-Caused Outages in Overhead Distribution Systems , 2006, 2006 International Conference on Probabilistic Methods Applied to Power Systems.

[2]  David Heckerman,et al.  Causal independence for probability assessment and inference using Bayesian networks , 1996, IEEE Trans. Syst. Man Cybern. Part A.

[3]  Hong Yu,et al.  An application of online learning algorithm for Bayesian network parameter , 2003, Proceedings of the 2003 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.03EX693).

[4]  Kevin Murphy,et al.  A brief introduction to graphical models and bayesian networks , 1998 .

[5]  Min Gui,et al.  Analysis of Animal-Related Outages in Overhead Distribution Systems With Wavelet Decomposition and Immune Systems-Based Neural Networks , 2009, IEEE Transactions on Power Systems.

[6]  Shao-Zhong Zhang,et al.  Construction and application of Bayesian networks in flood decision supporting system , 2002, Proceedings. International Conference on Machine Learning and Cybernetics.