Estimation of line flows and bus voltages using decision trees

A machine learning method called the ID3 (Interative Dichotomizer 3) approach is presented for the estimation of line flows and bus voltages following an outage event. A decision tree which is capable of generating the desired line flows and bus voltages are created using the training patterns which are compiled from the historical operating records of Taiwan power system. The established decision tree contains the knowledge which is essential for line flow and bus voltage prediction. Thus, it can be applied to estimate line flows and bus voltages of a system in an efficient manner. The effectiveness of the proposed ID3 approach is demonstrated by security assessment of Taiwan power system which contains 170 buses and 207 lines