Experimental evaluation of differential voltage protection scheme based on a coherence function applied to AC machine...

Area of Science:

• Electrical Engineering
• Power Systems Protection
• Machine Diagnostics

Background:

• Induction machines are critical in industrial applications.
• Stator winding faults can lead to significant downtime and operational costs.
• Existing fault detection methods may lack precision or speed.

Purpose of the Study:

• To introduce a coherence algorithm for detecting and classifying stator winding faults in induction machines.
• To differentiate between turn-to-turn and internal shunt faults.
• To evaluate the performance of the proposed fault detection technique.

Main Methods:

• Utilizing a coherence algorithm based on voltage measurements from stator winding terminals and midpoints.
• Quantifying cross-coherence and auto-coherence coefficients for fault analysis.
• Employing a modified induction motor with re-wounded stator windings for comprehensive testing.

Main Results:

• The coherence algorithm accurately detects and classifies various stator winding faults, including turn-to-turn and internal shunt types.
• Achieved dependability and security rates exceeding 98.0%, with reliability and accuracy near 97.0%.
• Demonstrated a high-speed, adjustable response with novel tripping curves.

Conclusions:

• The proposed coherence-based approach provides an effective and efficient solution for advanced induction machine protection.
• The technique reliably identifies fault types and locations within the protection zone.
• The method offers significant improvements in protection dependability, security, reliability, and accuracy.