Voltage Security Margin in Power Systems using Fuzzy Inference System and QV Sensitivity Analysis

D. Ceron, M. Rios, T. Herrera, and A. Torres (Columbia)


Continuation Power Flow, Fuzzy Inference Systems, Neuro-fuzzy networks, Voltage stability, QV analysis.


This paper presents a methodology to predict a security index of power systems for forecasted operating points based on a voltage security diagnostic and a modal analysis of QV sensitivity matrix in regard to the voltage stability modeling with fuzzy inference systems, like ANFIS. Local variables like active and reactive lines power flows are used as input variables and reduced by the technique known as principal component analysis (PCA). Two methods are used in order to find a security index for the inference process. The first method uses the loading parameter of the continuation power flow and the second uses a QV modal analysis criteria. These methodologies allow the calculation of the distance from the current operation point to the collapse point taking into account different possible conditions in a day ahead operating planning. These methodologies are tested and validated with the RTS-96 single area and the computer times of each method are compared.

Important Links:

Go Back