Deep Learning-Based Multimode Fiber Distributed Temperature Sensing

doi:10.3390/s25092811

Deep Learning-Based Multimode Fiber Distributed Temperature Sensing

2025-04-29

0

OA

AI

DOI：10.3390/s25092811

Original

Original request for help

Save

Abstract

En 中文

Highlights What are the main findings? We developed a fiber-optic temperature sensing method using Convolutional Neural Networks (CNNs). By inputting a speckle pattern into the CNN, we can determine the temperature at different locations of the fiber simultaneously; The network training was divided into three steps: first, training for temperature prediction; second, training for heating location prediction; and third, combined training for both temperature and heating location using all datasets; We tested the model with two types of optical fibers and achieved a satisfactory prediction accuracy. What are the implications of the main findings? This method addresses the high cost, complex installation, and limited accuracy of traditional fiber-optic temperature sensing technologies. It simplifies the measurement process, bypassing the need for complex physical models and offering a new efficient solution for fiber-optic sensing; Suitable for hazardous environments and other complex scenarios, this method allows the accurate acquisition of temperature and location information without direct contact with the target object.Highlights What are the main findings? We developed a fiber-optic temperature sensing method using Convolutional Neural Networks (CNNs). By inputting a speckle pattern into the CNN, we can determine the temperature at different locations of the fiber simultaneously; The network training was divided into three steps: first, training for temperature prediction; second, training for heating location prediction; and third, combined training for both temperature and heating location using all datasets; We tested the model with two types of optical fibers and achieved a satisfactory prediction accuracy. What are the implications of the main findings? This method addresses the high cost, complex installation, and limited accuracy of traditional fiber-optic temperature sensing technologies. It simplifies the measurement process, bypassing the need for complex physical models and offering a new efficient solution for fiber-optic sensing; Suitable for hazardous environments and other complex scenarios, this method allows the accurate acquisition of temperature and location information without direct contact with the target object.Abstract As a laser beam passes through a multimode fiber (MMF), a speckle pattern is generated, which is sensitive to temperature, thereby making the MMF a temperature-sensing element. A deep learning technique is employed to the MMF-based temperature sensor, to obtain high-precision temperature sensing. We designed an MMF-based temperature-sensing configuration and developed a dual-output Convolutional Neural Network (CNN) for predicting both the temperature and the position of the heating point, and we constructed a dataset. It was shown that the location prediction accuracy reached 100%, while the temperature prediction accuracy (within a +/- 1 degrees C error margin) was 100% and 95.12% in the two experiments, respectively. The precision of the predicting heating point was less than 1 cm. Different types of MMFs were used in temperature measurements, showing that the accuracy remained quite high. This non-contact, high-precision MMF-based temperature measurement method, driven by deep learning, is suitable for applications in hazardous environments.

Keywords:

convolutional neural networks

multimode fibers

temperature prediction

position prediction

speckle imaging

distributed sensing

Deep Learning-Based Multimode Fiber Distributed Temperature Sensing

Abstract

Journal

Organization

Cited Papers