首页|New Findings on Machine Learning from National University of Defense Technology Summarized (Dual-parametric Simultaneous Demodulation of Fiber Optic Seawater Temperature and Pressure Sensors Based On Machine Learning Methods)
New Findings on Machine Learning from National University of Defense Technology Summarized (Dual-parametric Simultaneous Demodulation of Fiber Optic Seawater Temperature and Pressure Sensors Based On Machine Learning Methods)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
IEEE
Data detailed on Machine Learning have been presented. According to news reporting originating from Changsha, People’s Republic of China, by NewsRx correspondents, research stated, “The pressure and temperature of seawater are two important parameters. At present, people mainly rely on various types of temperature and depth measurement instruments to monitor the temperature and pressure of seawater.” Financial supporters for this research include National Natural Science Foundation of China (NSFC), Interdisciplinary Scientific Research Foundation of Guangxi University, National Natural Science Foundation of Guangxi Province, Guangxi Major Projects of Science and Technology, Guangxi Key Projects of Science and Technology. Our news editors obtained a quote from the research from the National University of Defense Technology, “Optical temperature-depth (TD) sensors have advantages such as antielectromagnetic interference, corrosion resistance, and multiplexing capability. By using polydimethylsiloxane with a high thermo-optical coefficient and a high elastic-optical coefficient to encapsulate optical microfiber coupler combined sagnac loop (OMCSL) structure with large abrupt field characteristic, a simultaneous temperature and pressure fiber optic sensor with high stability and high sensitivity could be achieved. However, when using the conventional sensitivity matrix method (SMM) to demodulate the sensor, the demodulation results were unstable and encountered large error. One of the main reasons for the errors in the demodulation of the sensor using SMM is that the sensitivity matrix is an ill-conditioned matrix under certain conditions, and SMM in this state would greatly amplify the errors in the demodulation results. The other reason is that the feature wavelengths of the sensor would show a nonlinear relationship with temperature when sensing in some environments. To reduce the demodulation error, in this article, we researched and used various machine learning methods (MLMs) to demodulate the sensor. The experimental results showed that the demodulation error of the sensor could be greatly reduced by using the MLM compared to the traditional SMM.”
ChangshaPeople’s Republic of ChinaAsiaCyborgsEmerging TechnologiesMachine LearningNational University of Defense Technology
NETL
NSTL
IEEE
