Distance Relay Fault Classification for IBR-Based Power Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The integration of inverter-based resources (IBRs) in power grids leads to reduced fault current levels and instability, posing challenges for conventional protection schemes, including mis-operation or non-operation of distance relays due to low fault currents and lack of negative quantities, necessitating improved techniques for phase and zone classification and protection.
Innovation Solution
An AI/ML-based method utilizing a sliding window moving average and standard deviation to determine fault inception time, combined with random forest models for phase and zone classification, providing autonomous and self-setting protection functions for power systems with high IBR penetration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional protection schemes are used in power systems with high IBR penetration, then the protection system structure remains simple and well-established, but the reliability of fault detection deteriorates due to reduced fault current levels and stability margins
Solution Approach 1:
The patent changes the parameters used for fault detection from conventional current magnitude-based parameters to rate of change of current (di/dt) and voltage (du/dt) parameters. This allows the protection scheme to remain reliable in IBR-based systems where fault current levels are reduced, as the rate of change parameters maintain their discriminative power even when absolute current levels are lower.
Solution Approach 2:
The patent replaces conventional phasor-based distance protection algorithms with an AI/ML-based classification system. This substitution enables the protection scheme to adapt to IBR-based power systems by using learned patterns from training data, providing both reliability and adaptability where conventional methods fail.
2Reliability
If AI/ML-based classification systems are implemented for fault detection in IBR-based power systems, then the reliability and adaptability of protection improve, but the device complexity and computational requirements increase
Solution Approach 1:
The patent segments the AI/ML-based protection system into distinct functional modules: feature extraction module that computes rate of change parameters, training module that learns from labeled data, and classification module that performs fault detection. This segmentation manages complexity by organizing the system into manageable, well-defined components with clear interfaces.
Solution Approach 2:
The patent implements a preliminary training phase where the AI/ML model is trained offline using labeled training data from various fault scenarios. This preliminary action prepares the classification system in advance, so that during actual operation, only inference is required, reducing real-time computational complexity while maintaining high accuracy.
3Speed
If rate of change parameters are used for fault classification, then the speed of fault detection improves by enabling faster classification, but the measurement precision requirements increase
Solution Approach 1:
The patent implements continuous monitoring of voltage and current signals to compute rate of change parameters. By maintaining continuous measurement and computation, the system achieves both high speed detection (as faults are detected immediately when rate of change thresholds are exceeded) and adequate measurement precision (through continuous data collection that averages out noise).
Data Source
AI summary
Techniques for distance protection of a transmission line include determining a fault inception time from a voltage and/or current, determining rate of change sample values indicative of a rate of change of the voltage and/or of a rate of change of the current for at least one sample time that is dependent on the fault inception time, and using the rate of change sample values to generate a phase classifier for fault classification of a zone classifier for faulted zone identification.