Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Real-Time Ground Fault Detection for Inverter-Based Microgrid Systems

Jingwei Dong, Yucheng Liao, Haiwei Xie, Jochen Cremer, Peyman Mohajerin Esfahani

TL;DR

Real-time ground fault detection in inverter-based microgrids is challenging due to partial disturbance decoupling and modeling uncertainties. The authors propose a data-assisted residual-generation scheme within a differential-algebraic-equation (DAE) framework, cast as a quadratic-programming (QP) problem to design a fault-detection filter that uses only the current measurements $i_{odq}$ and robustly rejects disturbances $d$ and modeling error $oldsymbol{\xi}$. They provide an approximate analytical solution and a probabilistic false-alarm threshold via Markov inequality, enabling online implementation. Validation with high-fidelity simulations in Simulink and RTDS demonstrates reliable fault detection under disturbance and uncertainty while avoiding false alarms from load changes, highlighting practical potential for sensor-efficient, real-time IBM protection.

Abstract

Ground fault detection in inverter-based microgrid (IBM) systems is challenging, particularly in a real-time setting, as the fault current deviates slightly from the nominal value. This difficulty is reinforced when there are partially decoupled disturbances and modeling uncertainties. The conventional solution of installing more relays to obtain additional measurements is costly and also increases the complexity of the system. In this paper, we propose a data-assisted diagnosis scheme based on an optimization-based fault detection filter with the output current as the only measurement. Modeling the microgrid dynamics and the diagnosis filter, we formulate the filter design as a quadratic programming (QP) problem that accounts for decoupling partial disturbances, robustness to non-decoupled disturbances and modeling uncertainties by training with data, and ensuring fault sensitivity simultaneously. To ease the computational effort, we also provide an approximate but analytical solution to this QP. Additionally, we use classical statistical results to provide a thresholding mechanism that enjoys probabilistic false-alarm guarantees. Finally, we implement the IBM system with Simulink and Real Time Digital Simulator (RTDS) to verify the effectiveness of the proposed method through simulations.

Real-Time Ground Fault Detection for Inverter-Based Microgrid Systems

TL;DR

Real-time ground fault detection in inverter-based microgrids is challenging due to partial disturbance decoupling and modeling uncertainties. The authors propose a data-assisted residual-generation scheme within a differential-algebraic-equation (DAE) framework, cast as a quadratic-programming (QP) problem to design a fault-detection filter that uses only the current measurements and robustly rejects disturbances and modeling error . They provide an approximate analytical solution and a probabilistic false-alarm threshold via Markov inequality, enabling online implementation. Validation with high-fidelity simulations in Simulink and RTDS demonstrates reliable fault detection under disturbance and uncertainty while avoiding false alarms from load changes, highlighting practical potential for sensor-efficient, real-time IBM protection.

Abstract

Ground fault detection in inverter-based microgrid (IBM) systems is challenging, particularly in a real-time setting, as the fault current deviates slightly from the nominal value. This difficulty is reinforced when there are partially decoupled disturbances and modeling uncertainties. The conventional solution of installing more relays to obtain additional measurements is costly and also increases the complexity of the system. In this paper, we propose a data-assisted diagnosis scheme based on an optimization-based fault detection filter with the output current as the only measurement. Modeling the microgrid dynamics and the diagnosis filter, we formulate the filter design as a quadratic programming (QP) problem that accounts for decoupling partial disturbances, robustness to non-decoupled disturbances and modeling uncertainties by training with data, and ensuring fault sensitivity simultaneously. To ease the computational effort, we also provide an approximate but analytical solution to this QP. Additionally, we use classical statistical results to provide a thresholding mechanism that enjoys probabilistic false-alarm guarantees. Finally, we implement the IBM system with Simulink and Real Time Digital Simulator (RTDS) to verify the effectiveness of the proposed method through simulations.
Paper Structure (12 sections, 3 theorems, 51 equations, 6 figures)

This paper contains 12 sections, 3 theorems, 51 equations, 6 figures.

Table of Contents

  1. Introduction
  2. Modeling and Problem Statement
  3. System description
  4. State-space model of the fault-free IBM system
  5. State-space model of the IBM system with ground faults
  6. Problem statement
  7. Main Results
  8. Filter design
  9. Numerical results in RTDS
  10. Conclusions
  11. Simulation results for the perfect setting
  12. Comparison with other approaches

Key Result

Theorem 3.1

Consider the unified state-space model of the IBM system eq:ss4unifiedmodel and the structure of the fault detection filter in eq:Filter. Given the degree $d_N$, a stable $a(\mathfrak{q})$, and $m$ instances of output discrepancies $\xi_i$ and non-decoupled disturbances $\check{d}_{i}$, conditions e where $\bar{\Phi} = \frac{1}{m}\sum^{m}_{i=1} \Phi_i$, $\bar{\Psi} = \frac{1}{m}\sum^{m}_{i=1} \Psi

Figures (6)

  • Figure 1: Architecture of an IBM system with the diagnosis component.
  • Figure 2: Output currents generated by different models.
  • Figure 3: Diagnosis results with different models.
  • Figure 4: Diagnosis results for the perfect setting.
  • Figure 5: Diagnosis results using wavelet transform analysis.
  • ...and 1 more figures

Theorems & Definitions (13)

  • Remark 2.1: Disturbance decoupling condition
  • Remark 2.2: Discretization
  • Theorem 3.1: Filter design: QP
  • proof
  • Remark 3.2: Feasibility analysis
  • Remark 3.3: Perfect setting
  • Corollary 3.4: Approximate analytical solution
  • proof
  • Remark 3.5: Average objective function
  • Remark 3.6: Approximate analytical solution with $\delta$
  • ...and 3 more