Evolving fault diagnosis scheme for unbalanced distribution network using fast normalized cross-correlation technique

Area of Science:

• Electrical Engineering
• Power Systems Analysis
• Fault Detection and Diagnosis

Background:

• Unbalanced distribution systems face challenges in identifying and locating evolving faults.
• Existing solutions are often unsatisfactory for dynamic fault scenarios.
• Evolving faults involve sequential fault events with phase changes.

Purpose of the Study:

• To develop a robust method for identifying and locating evolving faults in unbalanced distribution systems.
• To differentiate between various types of evolving faults, including short circuits and cross-country faults.
• To assess the proposed method's effectiveness using real-time simulations.

Main Methods:

• Utilizing the cross-correlation technique for fault detection and location.
• Employing monitored network signals to identify the occurrence of the second fault (QRS value).
• Simulating evolving faults like short-to-short and cross-country faults.
• Applying the slime mold optimization approach to determine optimal monitoring points.

Main Results:

• The proposed cross-correlation technique accurately differentiates between various evolving fault types.
• Simulations on the IEEE 240 bus system demonstrate the method's capability to detect changing faults.
• The approach is invariant to fault characteristics (location, resistance, inception angle), monitoring point, and sample frequency.

Conclusions:

• The developed technique provides a satisfactory solution for identifying and locating evolving faults in unbalanced distribution systems.
• The method offers high accuracy and robustness across diverse fault conditions and system parameters.
• Optimal monitoring point selection using slime mold optimization enhances fault localization accuracy.