FPGA-Based Smart Sensor to Detect Current Transformer Saturation during Inrush Current Measurement.

Area of Science:

• Electrical Engineering
• Power Systems
• Measurement Science

Background:

• Current transformer (CT) saturation significantly degrades measurement accuracy and compromises power system protection reliability.
• Distinguishing between fault conditions and transformer inrush currents is a critical challenge for overcurrent protection systems.
• Existing methods for CT saturation detection often struggle with the unique waveforms of inrush currents.

Purpose of the Study:

• To develop and validate an FPGA-based smart sensor for accurate detection of current transformer saturation.
• To address the limitations of current methods, particularly their ineffectiveness during transformer inrush current events.
• To provide a robust and efficient solution for enhancing the reliability of power system protection.

Main Methods:

• Implementation of a novel algorithm on an FPGA-based smart sensor.
• Utilizing two time-domain features of the secondary current: the second-order difference function and the third-order statistic central moment.
• Testing the smart sensor under various conditions, including different residual flux, resistive burdens, sampling frequencies, and noise levels, as well as real-time experimental validation with an anti-aliasing filter.

Main Results:

• The proposed smart sensor achieves approximately 100% accuracy in detecting the start of CT saturation.
• The end of saturation is detected with a maximum error of a single sample.
• The system demonstrates high effectiveness, noise immunity, and accurate performance across diverse operating conditions, including challenging inrush current scenarios.
• Real-time experimental validation confirms the sensor's accurate results and efficient operation.

Conclusions:

• The developed FPGA-based smart sensor effectively detects current transformer saturation, even during inrush currents.
• The sensor offers high accuracy, robustness against noise, and reliable performance, surpassing existing methods.
• Its high-speed processing, simplicity, and low computational cost make it a valuable tool for improving power system protection reliability.