Classification of power system stability using support vector machines

The last years´ blackouts have indicated that even when a lot of data is available, the operators at different centers do not take the proper actions in time. This depends partly on the reorganization of the centers after the deregulation and partly on the lack of reliable supportive applications when the system is close to instability. This paper uses a novel technique based on support vector machines, SVM, in order to classify, if the power system can withstand a (n-1)-fault during a variety of operational conditions. The support vectors can be used on-line in order to determine if the system is moving into dangerous conditions and support the operators on an early stage, so proper actions can be made. This paper also shows that the scaling of the variables is important for good results. A new technique for finding the most important variables to measure or supervise is also presented.