Security Evaluation and Preventive Control of Power System using Decision Tree

摘要

This paper presents a decision tree based method for on line security evaluation and preventive control of power system. The method uses CART algorithms for generating the decision tree classifier using an off line data set. The classifier is then used online for security evaluation and preventive control purposes. In case an operating state is found to be insecure operating state, the preventive control strategy as obtained by the classifier is implemented.. The classifier gives two type of preventive control strategy namely generation rescheduling and optimal load shading. The choice of the preventive control strategy would depend upon the feasibility and economy of the power system operation. The method has been applied on an IEEE test systems and the results have been reported. The effect of various design parameters on the performance of the classifier has also been investigated. The proposed method has also been compared with the other Method.