A nonlinear fuzzy assisted image reconstruction algorithm for electrical capacitance tomography

Area of Science:

• Tomography
• Process Monitoring
• Materials Science

Background:

• Electrical Capacitance Tomography (ECT) is crucial for process monitoring.
• Accurate initial imaging is vital for iterative reconstruction convergence in ECT.
• Lost Foam Casting (LFC) requires precise molten metal characterization.

Purpose of the Study:

• To propose a nonlinear Fuzzy Inference System (FIS) method to enhance ECT image quality.
• To apply FIS for molten metal characteristic estimation in LFC.
• To improve the accuracy and convergence rate of ECT image reconstruction.

Main Methods:

• Development of a Fuzzy Inference System (FIS) to assess the likelihood of metal presence in the initial ECT image.
• Integration of the FIS-generated first image into an iterative image reconstruction algorithm.
• Utilizing eight capacitive sensors for metal position detection at the imaging area periphery.

Main Results:

• The proposed FIS method successfully computed the possibility of metal presence in the initial image.
• The FIS-enhanced first image accelerated the convergence rate of iterative reconstruction.
• The technique effectively detected metal position at the periphery using limited sensors.
• FIS demonstrated superior performance compared to standard iterative back projection.

Conclusions:

• The Fuzzy Inference System (FIS) offers a significant improvement for Electrical Capacitance Tomography (ECT) image quality in LFC.
• This approach enhances molten metal estimation accuracy and speeds up image reconstruction.
• The method provides a robust solution for process monitoring in LFC using ECT.