Core Concepts
This article argues for a next-generation grid monitoring system based on continuous point-on-wave measurements with generative AI for improved situational awareness and control.
Abstract
The article discusses the need for advanced grid monitoring systems due to changes in power grids, emphasizing the importance of high-resolution data. It introduces an innovative approach using continuous point-on-wave measurements and generative AI for fault detection and compression. The work is rooted in the Wiener-Kallianpur innovation representation theory, offering insights into anomaly detection and protection strategies. By comparing different methods, including conventional approaches and adaptive techniques, the study highlights the benefits of innovation-based sequential fault detection in reducing errors and improving decision time.
The research evaluates performance through simulations on an IEEE 13-bus distribution network under various fault scenarios. Results show that the proposed ISFD method outperforms conventional protection methods in terms of accuracy, false positive rates, and detection delays. The study provides detailed insights into test statistics under different methods, showcasing the effectiveness of innovation-based approaches for grid monitoring.
Stats
Outage events increased by 78% in the decade of 2011-2021 over the previous one.
A monitoring system based on CPOW data streaming should be fundamental for future grid situational awareness.
PMUs cover only a small fraction of significant grid events due to their limited reporting rates.
The rate of communications has increased by nearly six orders of magnitude since the invention of PMU technology.
Anomaly detection approach demonstrated 16% improvements in detection accuracy and 67.9% in detection speed.
Quotes
"don’t use those [other] methods—use a smooth test!" - Rayner and Best (1946)