| Abstract | 第1-10页 |
| 摘要 | 第10-16页 |
| Ⅰ.Introduction | 第16-32页 |
| 1 Introduction to Stationarity and Nonstationarity | 第16-19页 |
| ·Classical Definition of Stationarity | 第16-17页 |
| ·Generalized Notion of Stationarity | 第17-18页 |
| ·Issues in Representation of Nonstationary Signals | 第18-19页 |
| 2 Introduction to Time-Frequency Method | 第19-24页 |
| ·Fourier Transform | 第19-20页 |
| ·Characteristics of Time-Frequency Representation | 第20-21页 |
| ·Short-Time Fourier Transform | 第21页 |
| ·Wigner-Ville Distribution | 第21-23页 |
| ·Wigner-Ville Spectrum | 第23-24页 |
| 3 Reassignment | 第24-27页 |
| ·Spectrogram | 第24-25页 |
| ·Reassigned Spectrogram | 第25-27页 |
| ·Some Historical Comments | 第27页 |
| 4 Multitapering | 第27-32页 |
| ·Power Spectrum Density | 第27-28页 |
| ·Welch Method of Averaged Periodograms | 第28页 |
| ·Thomson's Method of Multitapering | 第28-30页 |
| ·Extensions of Multitapering | 第30-32页 |
| Ⅱ.Multitaper Time-Frequency Reassignment | 第32-60页 |
| 1 Multitaper Time-Frequency Reassignment for Nonstationary Spectrum Estimation | 第33-41页 |
| ·Principle and Implementation | 第33-36页 |
| ·Wedding multitapering with reassignment | 第33页 |
| ·Choice of tapers | 第33-34页 |
| ·Effectiveness measure by Renyi Entropy | 第34-36页 |
| ·Performance Evaluation:Error Measure | 第36-39页 |
| ·Examples | 第39-41页 |
| 2 Variations for Chirp Enhancement | 第41-49页 |
| ·Principle and Implementation | 第42-45页 |
| ·Differences between estimates based on different tapers | 第42-45页 |
| ·Final estimate combined by "sums" and "differences" | 第45页 |
| ·Performance Evaluations:Contrast Measure | 第45-47页 |
| ·Examples | 第47-49页 |
| 3 Application:Euler's Disk | 第49-59页 |
| ·Time-Frequency Estimation | 第51-52页 |
| ·Parameter Estimation of Chirp by Hough Transform | 第52-54页 |
| ·Performance Evaluation on Synthetic Model | 第54-57页 |
| ·Test on Experimental Signal | 第57-59页 |
| 4 Conclusion | 第59-60页 |
| Ⅲ.Test of Stationarity with Surrogates | 第60-94页 |
| 1 Time-Frequency Framework:Comparison of Local vs.Global Features | 第61-62页 |
| 2 Surrogates:Stationary Reference | 第62-65页 |
| ·Surrogates in(Non)stationarity Test | 第62-64页 |
| ·Surrogates in(Nonlinear)Physics | 第64-65页 |
| 3 Stationarity Test-Parametric Model Approach | 第65-82页 |
| ·Comparison of Local vs.Global Time-Frequency Spectra | 第65页 |
| ·Choice of Distance | 第65-67页 |
| ·Test signals | 第65-66页 |
| ·Probability laws distances | 第66-67页 |
| ·Spectral distances | 第67页 |
| ·Combined distance | 第67页 |
| ·Structure of the Test | 第67-69页 |
| ·Null Hypothesis of Stationarity | 第69-77页 |
| ·Distribution based on surrogates | 第69页 |
| ·Gamma model? | 第69-71页 |
| ·How many surrogates needed? | 第71-74页 |
| ·Threshold based on a statistical significance | 第74页 |
| ·Verification of null hypothesis | 第74-77页 |
| ·Index and Scale of Nonstationarity | 第77页 |
| ·Examples | 第77-82页 |
| 4 Stationarity Test-One-Class Support Vector Machine Approach | 第82-89页 |
| ·One-Class Support Vector Machines | 第82-85页 |
| ·Machine learning | 第82-83页 |
| ·Support Vector machines | 第83页 |
| ·One-class Support Vector machines | 第83-85页 |
| ·Comparison of Local vs.Global Time-Frequency Spectra | 第85-86页 |
| ·Examples | 第86-89页 |
| 5 Application:Speech | 第89-92页 |
| 6 Conclusion | 第92-94页 |
| Ⅳ.Conclusions and Perspectives | 第94-96页 |
| Ⅴ.Appendix | 第96-110页 |
| 1 Alternative model | 第96-99页 |
| 2 Choice of Thresholds | 第99-102页 |
| 3 Variations of Test | 第102-108页 |
| 4 Publications | 第108-110页 |
| Bibliography | 第110-116页 |
| Acknowledgment | 第116-117页 |
