An Intelligent Combination Method for Static Security Assessment

Power system security assessment plays a crucial role in maintaining safe and secure system operation. Conventional means of security analysis are based on the resolution of load-flow equations for each contingency to compute a performance index (PI). This technique seems to be unsatisfying due to the time-consuming and intensive computational efforts. In this paper, we present an approach based on the combination of Fuzzy C-Means clustering (FCM) algorithm and artificial neural network (ANN) to classifying the power system status as critical, secure and insecure for a certain operating condition and a specified contingency. The training set of the FCM-ANN method is created through offline simulations. Then, Direct- Method is used in the feature selection process to reduce the dimensionality of the inputs. The security classification is performed with the help of the PI computed using Newton-Raphson load flow method. The proposed method is tested on the New England 39-bus system. As compared to a multilayer feed forward network (MFFN), the FCM-ANN yields better results in terms of accuracy and rapidity.