A Novel Online Estimation Scheme for Static Voltage Stability Margin Based on Relationships Exploration in a Large Data Set

A novel integrated scheme based on relationships exploration (RE) is presented for the online estimation of the relative voltage stability margin (VSM). The proposed scheme can select the optimal variables as input features for estimation by detecting the relationship between each operation variable and the relative VSM in a large data set. Each relationship is given scores by the maximal information coefficient (MIC) and the Pearson correlation coefficient (PCC). The variables selected for estimation are corresponding to the relationships highly ranked by MIC and PCC, including linear relationships and nonlinear functional ones. Some of the highly ranked relationships are shown, curve fitted and described from the perspective of power system operation. If the data of the selected variables are obtained online from the wide area measurement system (WAMS), the relative VSM can be online estimated based on the explored relationships. The integrated scheme using RE-based process is examined on a 21-bus test system and a practical 1648-bus system provided by the software PSS/E, and it presents acceptable estimation accuracy. The impacts of training set size, selected relationships' types, total number and ranks on estimation accuracy are analyzed. The robustness of the scheme to measurement errors and topology variation is studied.