跳到主要內容

臺灣博碩士論文加值系統

(100.28.227.63) 您好!臺灣時間:2024/06/15 00:03
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:王昶文
研究生(外文):Chang-Wen Wang
論文名稱:運用小波轉換及神經網路在開迴路光纖陀螺儀系統上消除雜訊
論文名稱(外文):Elimination of Noise in Open-loop Fiber Optic Gyroscope by Wavelet Transform and Neural Network
指導教授:魏嘉建
指導教授(外文):Wei,Chia-Chien
學位類別:碩士
校院名稱:國立中山大學
系所名稱:光電工程學系研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2020
畢業學年度:108
語文別:中文
論文頁數:55
中文關鍵詞:開迴路光纖陀螺儀小波轉換神經網路小波降噪訊號處理
外文關鍵詞:wavelet transformopen-loop fiber optic gyroscopesignal processingneural networkwavelet denoising
相關次數:
  • 被引用被引用:0
  • 點閱點閱:134
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
干涉型光纖陀螺儀 (Interferometric Fiber-Optic Gyroscope, IFOG) 有著尺寸小、架構簡單等等優點,隨著技術發展,其靈敏度已經提升至足以在工業應用上實踐的程度。然而,即使是微小的雜訊,長久累積都可能會造成結果有相當大的誤差,因此雜訊的降低是使用光纖陀螺儀量測的一門重要課題。
本篇論文中針對開迴路光纖陀螺儀 (Open-loop Fiber Optic Gyroscope) 系統量測到的訊號,設計了弦波、平坦形狀的參考訊號作為訓練神經網路的參考訊號。接著我們利用深度神經網路、小波神經網路、混合式上升深度神經網路、混合式上升小波神經網路以及上升小波轉換,對實際量測的光纖陀螺儀訊號進行處理,以求降低非線性、非穩定特性的雜訊。我們有兩種指標:訊雜比 (Signal to Noise Ratio) 和角度隨機漫步 (Angle Random Walk) 來判斷訊號的品質。作為降噪效果的比較,我們在開迴路光纖陀螺儀的基礎上增加帶通濾波器限縮光源頻寬,以增加相對強度雜訊 (Relative Intensity Noise, RIN)。降噪處理過的訊號於特定架構下能提升0.4 dB的訊雜比,或是降低0.0003 deg/√hr 的角度隨機漫步,並對其可能成因進行了探討。
Interferometric Fiber-Optic Gyroscope (IFOG) has several advantages such as small size and simple structure. Thanks to the development of related technologies, the sensitivity of IFOG is high enough to achieve applications in industry. However, even the tiny noise may lead to lots of errors in long-term accumulation. Therefore, it’s an important task to reduce the noise in an IFOG.
In this work, we use a deep neural network, wavelet neural network, hybrid lifting deep neural network, hybrid lifting wavelet neural network or lifting wavelet transform to eliminate the noise in an open-loop IFOG. The reference signals in the training procedure were designed to be sinusoidal or flat. We have two indicators to determine the quality of signals. One is the signal to noise ratio (SNR) and the other one is the angle random walk (ARW) coefficient. As a comparison to the basic structure of open-loop FOG, we add a band-pass filter to reduce the bandwidth of the light source in order to increase its relative intensity noise (RIN). In some cases, the SNR can be improved by 0.4 dB, or the ARW coefficient can be reduced by 0.0003 deg/ √hr. Besides, related discussions are given to explain the experiment results.
論文審定書 i
致謝 ii
中文摘要 iii
Abstract iv
目錄 v
圖目錄 vii
表目錄 viii
第一章 緒論 1
1-1 前言 1
1-2 研究動機 2
第二章 干涉型光纖陀螺儀量測系統 3
2-1 干涉型光纖陀螺儀 3
2-1-1 薩格納克效應及基本原理 3
2-1-2 開迴路光纖陀螺儀 4
2-2 陀螺儀訊號的雜訊與飄移 7
2-2-1 雜訊、飄移與比例因子 7
2-2-2 艾倫方差 7
第三章 小波轉換以及神經網路之於訊號處理 10
3-1 小波轉換 10
3-1-1 小波轉換簡介 10
3-1-2 離散小波轉換與小波轉換上升方案 11
3-1-3 斯坦無偏風險估計決定臨界值降噪 14
3-2 神經網路 17
3-2-1 神經網路簡介 17
3-2-2 誤差反向傳播演算法與梯度下降法 19
3-2-3 基於神經網路之波形迴歸非線性補償 20
第四章 實驗結果與討論 22
4-1 實驗架構 22
4-1-1 實驗設備 22
4-1-2 訊號處理架構 25
4-2 實驗結果與討論 29
4-2-1 實際量測光纖陀螺儀訊號與參考訊號之設計 29
4-2-2 不同架構對於陀螺儀訊號處理結果之比較 35
4-2-3 問題與討論 41
結論 43
參考文獻 44
[1]G. T. Schmidt, “INS/GPS Technology Trends”, Low-Cost Navigation Sensors and Integration Technology, NATO RTO Lecture Series, RTO-EN-SET-116, March 2011.
[2]V. Vali and R. W. Shorthill, “Fiber Ring Interferometer”, Applied Optics, Vol. 15, Issue 5, pp. 1099-1100, 1976.
[3]Singh, B., Rana, H., Kumar, S., Bhulania, P., Minocha, G. “A Novel Design of Fiber Optic Gyroscope Based INS System for UAS Applications”, Procedia Comput. Sci. 57, pp. 1317-1323, 2015.
[4]A. Albrecht and J. Petereit. “Application of an Off-the-shelf Fiber Optic Gyroscope Based Inertial Measurement Unit for Attitude and Heading Estimation”, in 2017 IEEE SENSORS, pp. 1-3, Glasgow, UK, October 2017.
[5]J. Xie, L. Xu, E. Chen, “Image Denoising and Inpainting with Deep Neural Networks”, Adv. Neural Inf. Process. Syst. 25, pp. 350-358, 2012.
[6]Chang, S., Yu, B., Vetterli, M., “Adaptive Wavelet Thresholding for Image Denoising and Compression”, IEEE Trans. Image Process., Vol. 9, Issue 9, pp. 1532-1546, 2000.
[7]H. J. Arditty, H. C. Lefevre, "Sagnac Effect in Fiber Gyroscopes," Opt. Lett., Vol. 6, Issue 8, pp. 401-403, 1981.
[8]Ulrich, R., “Fiber-optic Rotation Sensing with Low Drift,” Opt. Lett., Vol. 5, pp. 173-175, 1980.
[9]“IEEE Standard Specification Format Guide and Test Procedure for Single-Axis Interferometric Fiber Optic Gyros”, IEEE Std. 952-1997, August 1998.
[10]S. Mallat, “A Theory for Multiresolution Signal Decomposition: The Wavelet Representation”, IEEE Trans. Pattern Anal. Mach. Intell., Vol. 11, pp. 674-693, July 1989.
[11] I. Daubechies, “Orthonormal Bases of Compactly Supported Wavelets”, Commun. Pure Appl. Math., Vol. 41, pp. 909-996, 1988.
[12]W. Sweldens, “The Lifting Scheme: A Custom-Design Construction Of Biorthogonal Wavelets”, Appl. Comput. Harmonic Analysis, Vol. 3, Issue. 2, pp. 186-200, 1996.
[13]D. L. Donoho, I. M. Johnstone, "Adapting to Unknown Smoothness by Wavelet Shrinkage", J. Amer. Statist. Assoc., 1995.
[14]M. Nielsen, “Neural Networks and Deep Learning”, [Online]. Available: http://neuralnetworksanddeeplearning.com/ (Accessed: December 2019)
[15]Q. Zhang, A. Benviste, “Wavelet Networks”, IEEE Trans. Neural Networks, Vol. 3, pp. 889-898, November 1992.
[16]S. Dang, K. Wang, H. Han, P. Cheng, “A Hybrid LWNN-Based Stochastic Noise Eliminating Method for Fiber Optic Gyro”, in 12th Int. Conf. Comput. Intell. Secur. (CIS), pp. 399-403, Wuxi, China, December 2016.
電子全文 電子全文(網際網路公開日期:20250209)
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊