Leveling Maintenance Mechanism by Using the Fabry-Perot Interferometer with Machine Learning Technology - Research findings

Aug 3, 2026

Upload by Finn

This study proposes a method for maintaining parallelism in a laser interferometer optical cavity by leveraging machine learning control. A Fabry-Perot interferometer serves as the experimental optical structure, enabling short optical paths and producing interference fringes linked to mirror parallelism. Supervised learning is trained with labeled interference images to classify and predict the tilt angle of a plane mirror. Stepper motors then adjust pitch and yaw automatically. Experiments report average correction errors and standard deviations of 32.38±11.21 arcseconds (17-grid) and 19.44±7.86 arcseconds (25-grid).

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages5

Reading Time13 min

AccessFree to download

Introduction

Theory and Principle

Fabry-Perot Interferometer

Machine Learning

ISSN 1846-6168 (Print), ISSN 1848-5588 (Online) Preliminary communication [https://doi.org/10.31803/tg-20230425154156](https://doi.org/10.31803/tg-20230425154156) Received: 2023-04-25, Accepted: 2023-04-28 Leveling Maintenance Mechanism by Using the Fabry-Perot Interferometer with Machine Learning Technology Syuan-Cheng Chang, Chung-Ping Chang*, Yung-Cheng Wang, Chi-Chieh Chu Abstract: This study proposes a method for maintaining parallelism of the optical cavity of a laser interferometer using machine learning. The Fabry-Perot interferometer is utilized as an experimental optical structure in this research due to its advantage of having a brief optical structure. The supervised machine learning method is used to train algorithms to accurately classify and predict the tilt angle of the plane mirror using labeled interference images. Based on the predicted results, stepper motors are fixed on a plane mirror that can automatically adjust the pitch and yaw angles. According to the experimental results, the average correction error and standard deviation in 17-grid classification experiment are 32.38 and 11.21 arcseconds, respectively. In 25-grid classification experiment, the average correction error and standard deviation are 19.44 and 7.86 arcseconds, respectively. The results show that this parallelism maintenance technology has essential for the semiconductor industry and precision positioning technology. Keywords: Fabry-Perot interferometer; interference image; leveling maintenance; machine learning; optical measurement 1 INTRODUCTION The precision machinery and semiconductor industries are critical to high-precision positioning technology, such as semiconductor production, μLED mass transfer, and wafer positioning processing, which demand extremely high positioning accuracy [1, 2] . As technology advances and human needs require smaller and more efficient products, positioning accuracy has increased from sub-micron to nanometer scale. Therefore, positioning technology is currently one of the most important and critical technologies in these industries [3, 4] . However, traditional requirements for straightness and parallelism are insufficient to meet the demand for higherprecision mechanical components. A better active leveling maintenance system is required to meet industry demands for parallel positioning correction. This research focuses on the development of a leveling maintenance mechanism (LMM) system that uses machine learning in conjunction with Fabry-Perot interferometer. The system is designed with considerations for optical structure, machine learning control, and feedback to enhance current industry technologies for precision parallelism correction. Based on the interference image, this study utilizes machine learning for training, effectively avoiding the accuracy and sensitivity issues caused by traditional methods. 2 THEORY AND PRINCIPLE To construct an active parallelism maintenance system using non-contact optical interference methods, we introduce the theory of the optical structure and the machine learning methods of LMM as follows. 2.1 Fabry-Perot Interferometer The Fabry-Perot interferometer (FPI) is an optical instrument consisting of two parallel reflecting mirrors that form a resonant cavity, as shown in Fig. 1. When light reflects inside the cavity, a series of interference fringes is formed, which are related to the parallelism of the resonant cavity [5]. When the parallelism of the two mirrors is better, the contrast and clarity of the interference fringes will be higher. Therefore, by observing the variation of interference fringes, we can determine the parallelism of the resonant cavity [6] . Figure 1 Fabry-Perot interferometer The parallelism of the resonant cavity can be determined by calculating the angle of reflection mirror displacement from the number of interference fringes, as shown in Fig. 2. The relationship between fringes and angles is direct: the an

FAQs

How does the Fabry-Perot interferometer indicate cavity parallelism in this system?

Interference fringes formed inside the resonant cavity relate directly to the parallelism of the two mirrors. Higher mirror parallelism yields clearer fringes, allowing the deviation to be inferred from fringe variation.

What machine learning task is used for estimating mirror tilt angles?

Supervised machine learning is applied in a classification setting using labeled interference images to predict the plane mirror’s tilt angle.

What accuracy results were achieved in the experiments?

In the 17-grid classification experiment, the average correction error and standard deviation are 32.38 and 11.21 arcseconds. In the 25-grid classification experiment, they are 19.44 and 7.86 arcseconds.

Aug 3, 2026

Upload by Finn