Loadability margin estimation of power system using model trees

This paper presents a model tree (MT) method for estimation of loadability margins of power system under single line outage contingency and varying loading conditions. Maximum loadability margin is used as an index for voltage security assessment. In this paper model trees estimate the loadability margins in terms of continuous values. Learning states for MT are generated by computer simulation using continuation power flow and limits on reactive generation are considered while calculating loadability margins. Once the model tree is built it acquires the knowledge on the basis of which it estimates post-contingent maximum loadability margins efficiently and within the desired accuracy. The computation speed of this method is very high (ultra-fast), hence it can be more useful for real time applications. The effectiveness of this approach is tested on IEEE 30-bus test system, using unknown test patterns.