Decision Tree-Based Online Voltage Security Assessment Using PMU Measurements

Voltage collapse is a critical problem that impacts power system operational security. Timely and accurate assessment of voltage security is necessary to detect post-contingency voltage problems in order to prevent a large scale blackout. This paper presents an online voltage security assessment scheme using synchronized phasor measurements and periodically updated decision trees (DTs). The DTs are first trained offline using detailed voltage security analysis conducted using the past representative and forecasted 24-h ahead operating conditions. The DTs are also updated every hour by including newly predicted system conditions for robustness improvement. The associated synchro-nized critical attributes are obtained in real time from phasor measurement units (PMUs) and compared with the offline thresh-olds determined by the DTs to assess security. This approach is tested on the American Electric Power (AEP) system and properly trained DTs perform well in assessing voltage security. Several new ideas to improve DT performance are also introduced