Displacement sensing in a multimode SNAP microcavity by an artificial neural network

Area of Science:

• Photonics
• Nanotechnology
• Machine Learning

Background:

• Surface Nanoscale Axial Photonics (SNAP) microcavities coupled with waveguides enable displacement sensing through axial modes.
• The relationship between coupling position and spectral modes is complex and nonlinear, hindering analytical modeling.

Purpose of the Study:

• To develop a Back-Propagation Neural Network (BPNN) model for the complex functional relationship in SNAP-based displacement sensing.
• To evaluate the potential of a multimode sensing scheme for practical, large-range, high-precision displacement sensing.

Main Methods:

• Utilizing the coupling between SNAP microcavities and waveguides.
• Applying a Back-Propagation Neural Network (BPNN) to model the nonlinear functional relationship between coupling position and transmission spectrum modes.
• Comparing the performance of the multimode sensing scheme with single-mode sensing using whispering gallery mode (WGM) resonators.

Main Results:

• The BPNN successfully models the complex, nonlinear functional relationship between coupling position and axial modes.
• The multimode sensing scheme demonstrates significant potential for large-range and high-precision displacement sensing applications.
• The SNAP-based multimode sensing shows advantages over traditional WGM single-mode sensing.

Conclusions:

• BPNN modeling provides an effective solution for the complex nonlinearities in SNAP-based displacement sensing.
• The SNAP multimode sensing scheme presents a promising platform for advanced, high-performance displacement measurement.
• This approach offers a viable alternative to existing sensing technologies, particularly for applications requiring large range and high precision.