ANN-Based Voltage Stability Assessment with PMU Infrastructure Placement Consideration

Artificial intelligence has led to increasingly significant technological advancements in a variety of sectors, including electrical power systems. The traditional methods to assess the stability of electric power systems typically need a significant amount of processing time and a complete set of network parameters. Many data points will be generated due to the high resolution of the Phasor Measurement Unit (PMU) measurement results. As a result, an efficient and effective way to process this data is required. The objective of this study is to establish a voltage stability detection method utilizing the artificial neural network approach by assessing its performance based on metrics. In addition, the costs associated with PMU and communication infrastructure will be assessed for each scenario. The imbalance in data distribution between classes and the location of PMU used in the dataset affects the performance of the detection. This research indicates that selecting an unsuitable PMU location can lead to poor performance in voltage stability assessment below 80. Furthermore, the choice of metrics should align with the characteristics and objectives of the analysis. It is also important to consider the cost variables to minimize the costs related to PMU installation while still ensuring the reliability of detection performance. © 2024 IEEE.