Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enhanced Entropy-Based Metric for Characterization of Delayed Voltage Recovery

Mohammad Almomani, Muhammad Sarwar, Venkataramana Ajjarapu

TL;DR

The paper tackles delayed voltage recovery (FIDVR) and the shortcomings of entropy-based KL divergence in oscillatory and over-voltage conditions. It proposes Enhanced Voltage Recovery Violation Index (EVRVI), which uses Empirical Mode Decomposition to extract monotonic upper and lower envelopes from voltage signals and applies KL divergence to quantify recovery for over- and under-voltage events. In Nordic-system simulations with about 245k scenarios, EVRVI outperforms the traditional KL approach by eliminating false negatives and providing robust OV/UV detection, albeit with a modest increase in computation time due to EMD. The work suggests EVRVI as a practical, more reliable tool for real-time FIDVR assessment and system monitoring, with future work focused on computational efficiency for large-scale deployment.

Abstract

Ensuring accurate violation detection in power systems is paramount for operational reliability. This paper introduces an enhanced voltage recovery violation index (EVRVI), a comprehensive index designed to quantify fault-induced delayed voltage recovery (FIDVR). EVRVI enhances traditional entropy-based methods by leveraging Empirical Mode Decomposition (EMD) to extract key features from the voltage signal, which are then used to quantify over-voltage (OV) and under-voltage (UV) events. Our simulations on the Nordic system, involving over 245k scenarios, demonstrate EVRVI's superior ability to identify and categorize voltage recovery issues compared to the traditional entropy-based measure. EVRVI not only significantly reduces false negatives in violation detection but also provides a reliable framework for over-voltage detection, making it an invaluable tool for modern power system studies.

Enhanced Entropy-Based Metric for Characterization of Delayed Voltage Recovery

TL;DR

The paper tackles delayed voltage recovery (FIDVR) and the shortcomings of entropy-based KL divergence in oscillatory and over-voltage conditions. It proposes Enhanced Voltage Recovery Violation Index (EVRVI), which uses Empirical Mode Decomposition to extract monotonic upper and lower envelopes from voltage signals and applies KL divergence to quantify recovery for over- and under-voltage events. In Nordic-system simulations with about 245k scenarios, EVRVI outperforms the traditional KL approach by eliminating false negatives and providing robust OV/UV detection, albeit with a modest increase in computation time due to EMD. The work suggests EVRVI as a practical, more reliable tool for real-time FIDVR assessment and system monitoring, with future work focused on computational efficiency for large-scale deployment.

Abstract

Ensuring accurate violation detection in power systems is paramount for operational reliability. This paper introduces an enhanced voltage recovery violation index (EVRVI), a comprehensive index designed to quantify fault-induced delayed voltage recovery (FIDVR). EVRVI enhances traditional entropy-based methods by leveraging Empirical Mode Decomposition (EMD) to extract key features from the voltage signal, which are then used to quantify over-voltage (OV) and under-voltage (UV) events. Our simulations on the Nordic system, involving over 245k scenarios, demonstrate EVRVI's superior ability to identify and categorize voltage recovery issues compared to the traditional entropy-based measure. EVRVI not only significantly reduces false negatives in violation detection but also provides a reliable framework for over-voltage detection, making it an invaluable tool for modern power system studies.

Paper Structure

This paper contains 15 sections, 6 equations, 2 figures.

Table of Contents

  1. Introduction
  2. Literature Review and Research Gap
  3. Slope-based methods:
  4. Integral error-based methods:
  5. Contribution of This Paper
  6. Over-voltage and Under-voltage Detection:
  7. Managing Oscillations During Recovery:
  8. Enhanced Comparison of Voltage Signals:
  9. Background
  10. KL Divergence Measure
  11. Empirical Mode Decomposition
  12. Enhanced Index Using EMD
  13. Step-by-Step Implementation
  14. Simulation Results
  15. conclusion

Figures (2)

  • Figure 1: False Detection cases using KL Divergence
  • Figure 2: Comprehensive analysis of delayed voltage recovery signals using the proposed extended index. The figure illustrates two cases: under-voltage (top row) and over-voltage (bottom row). The columns depict (1) voltage signals and stepwise reference criteria, (2) lower and upper envelopes calculated using equations (2) and (3), (3) probability distributions of the envelopes for $N$ segments, and (4) the EVRVI calculated using equations (5) and (6).