Long-term voltage stability monitoring of power system areas using a kernel extreme learning machine approach

Abstract

The ever increasing active and reactive power demands, along with limited sources of generation and delays in transmission expansion projects, have led many power systems to operate near their voltage stability limits. In this context, voltage stability monitoring methodologies have become an important topic in power systems research. This paper presents a novel methodology for long-term voltage stability monitoring in power systems that exploits the feasibility of phasor-type information in order to estimate the long-term voltage stability status. The information regarding the current system condition is acquired through synchronized phasor measurements and the power system is divided in sub-areas for improving its supervision; then, an artificial intelligence approach based on kernel extreme learning machine is used for long-term voltage stability assessment. The proposed scheme allows foreseeing the voltage instability caused by limitations in reactive power transmission, and it also permits alerting when a system area experiences a deficit of reactive power from supply sources. The validation of the proposed method is performed on the 39-bus test system, obtaining feasible results. The tests confirmed that the proposed method works properly under different scenarios and system conditions, always ensuring proper voltage stability status results independently of its cause.

Keywords

• Angle cut-set
• Kernel extreme learning machine (KELM)
• Long-term voltage stability monitoring
• Near real-time monitoring
• Voltage control area