Robust identification of multiple-input single-output system response for efficient pickup noise removal from tokamak...

Area of Science:

• Plasma physics
• Fusion energy research
• Signal processing

Background:

• Tokamak fusion devices generate significant electromagnetic noise.
• Weak diagnostic photocurrents (nA range) are susceptible to contamination from high-current coil crosstalk (kA range).
• Accurate measurement of diagnostic signals is crucial for fusion energy research.

Purpose of the Study:

• To develop a robust algorithm for identifying and mitigating electromagnetic pickup noise in tokamak diagnostic systems.
• To enable precise subtraction of crosstalk noise affecting weak photocurrent measurements.
• To improve the signal-to-noise ratio in diagnostic data.

Main Methods:

• Development of a linear multi-input single-output (MISO) system identification algorithm.
• Modeling the dynamic relationship between power source signals and diagnostic signals.
• Application of the algorithm to experimental data from the DIII-D tokamak.

Main Results:

• Successfully identified and characterized the MISO system representing noise pathways.
• Achieved significant noise reduction of up to 20 dB within the 1-20 kHz frequency range.
• Demonstrated the algorithm's effectiveness in real-world tokamak experimental conditions.

Conclusions:

• The developed MISO system identification algorithm provides a robust solution for electromagnetic noise reduction in tokamaks.
• This method enhances the accuracy of weak photocurrent measurements, crucial for fusion plasma diagnostics.
• The technique shows promise for improving data quality in future fusion energy experiments.