PMU-Based Voltage Stability Measurement Under Contingency Using ANN

Abstract

An artificial neural network (ANN) is an effective and efficient approach for solving problems, which require fast and complicated computation. The study proposed a PMU measurement-based voltage stability monitoring and assessment using ANN hybrid model for the evaluation of voltage stability margin. Particle swarm optimization (PSO) optimizes the meta-parameter of ANNs for improving its convergence rate. This model helps in real-time monitoring of long-term voltage stability to avoid grid blackout. The power system security is generally defined based on the single contingency “N−1” criterion, meaning normal system minus one element. Separate ANNs are trained and designed for selected worst case “N−1” contingencies along with base case configuration. Phasor Measurement Unit installed at distant locations provide real-time synchronized measurements of voltage phasor, which forms the input feature for the proposed model. Load <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$P$ </tex-math></inline-formula> margin evaluated by continuation power flow (CPF) method is considered as output feature of the proposed ANN model. The performance indices mean absolute percentage error (MAPE) and maximum percentage error (MPE) with smaller values indicate small deviation between the estimated and the actual values. The proposed approach is effectively tested on IEEE 14-bus and 30-bus test system.