| 摘要 | 第5-6页 |
| ABSTRACT | 第6-7页 |
| Chapter 1 Introduction | 第28-34页 |
| 1.1 Introduction and Motivation | 第28-30页 |
| 1.2 List of Contributions | 第30-31页 |
| 1.3 Thesis Organization | 第31-34页 |
| Chapter 2 Literature Review | 第34-52页 |
| 2.1 Artificial Neural Network | 第34-40页 |
| 2.1.1 Neural Network Structures | 第35-38页 |
| 2.1.2 Neural network Training | 第38-40页 |
| 2.2 Knowledge-Based Neural Network | 第40-41页 |
| 2.3 Space Mapping | 第41-48页 |
| 2.3.1 Space Mapping Concept | 第42-43页 |
| 2.3.2 Input Space Mapping | 第43-44页 |
| 2.3.3 Implicit Space Mapping | 第44-45页 |
| 2.3.4 Output Space Mapping | 第45-46页 |
| 2.3.5 Tuning Space Mapping | 第46页 |
| 2.3.6 Neuro Space Mapping | 第46页 |
| 2.3.7 Parallel Space Mapping | 第46-48页 |
| 2.4 Overview of Parametric Modeling and Optimization withNo Available Coarse Model | 第48-51页 |
| 2.4.1 Coarse and Fine Mesh Space Mapping | 第49页 |
| 2.4.2 Cognition-Driven Formulation of Space Mapping for Equal-RippleOptimization of Microwave Filters | 第49-50页 |
| 2.4.3 Parallel EM Optimization using Transfer Function-based Surro-gate Model | 第50页 |
| 2.4.4 Parametric Modeling using Combined Neural Networks and Trans-fer Functions | 第50-51页 |
| 2.5 Conclusion | 第51-52页 |
| Chapter 3 Parametric Modeling Using Pole-Residue-BasedNeuro-Transfer Functions | 第52-92页 |
| 3.1 Introduction | 第52-54页 |
| 3.2 Proposed Technique for Developing a Pole-Residue-BasedNeuro-TF Model | 第54-65页 |
| 3.2.1 Formulation of The Pole-Residue-Based Neuro-TF Model .. · 27 | 第54-56页 |
| 3.2.2 The Issue of Discontinuity in Poles and Residues with respect toGeometrical Parameters | 第56-58页 |
| 3.2.3 Vector Fitting for Parameter Extraction | 第58-59页 |
| 3.2.4 Pole-Residue Tracking Technique for Order-Changing | 第59-63页 |
| 3.2.5 Preliminary Training of Neural Networks | 第63-64页 |
| 3.2.6 Refinement Training of the Pole-Residue-Based Neuro-TF Model | 第64-65页 |
| 3.3 Application Examples | 第65-90页 |
| 3.3.1 Parametric Modeling of a Microwave Junction | 第65-74页 |
| 3.3.2 Parametric Modeling of an Ultra-Wideband Antenna | 第74-82页 |
| 3.3.3 Parametric Modeling of a Coupled-Line Filter | 第82-90页 |
| 3.4 Conclusion | 第90-92页 |
| Chapter 4 Advanced Pole-Residue Tracking Technique Us-ing EM Sensitivity Analysis | 第92-104页 |
| 4.1 Introduction | 第92-93页 |
| 4.2 Advanced Pole-Residue Tracking Technique Using EM Sen-sitivity Analysis | 第93-100页 |
| 4.2.1 Pole-Residue Extraction | 第94-96页 |
| 4.2.2 Advanced Pole-Residue Tracking Technique Using Derivative In-formation | 第96-100页 |
| 4.3 Application Example | 第100-102页 |
| 4.3.1 Illustration of Pole-Residue Tracking Technique Using a Coupled-Line Filter Example | 第100-102页 |
| 4.3.2 Advanced Pole-Residue Tracking Technique Using Sensitivity Anal-ysis of The Coupled-Line Filter Example | 第102页 |
| 4.4 Conclusion | 第102-104页 |
| Chapter 5 Parametric Modeling Using Neuo-Transfer Func-tions with EM Sensitivity Analysis | 第104-140页 |
| 5.1 Introduction | 第104-105页 |
| 5.2 Proposed Sensitivity-Analysis-Based Neuro-TF ModelingTechnique | 第105-117页 |
| 5.2.1 Pole-Residue Extraction and Derivative Data Calculation | 第107-111页 |
| 5.2.2 Preliminary Training of the Sensitivity-Analysis-Based Neuro-TFModel | 第111-112页 |
| 5.2.3 Refinement Training of the Sensitivity-Analysis-Based Neuro-TFModel | 第112-113页 |
| 5.2.4 Derivative Calculation for Training the Proposed Sensitivity-Analysis-Based Neuro-TF Model | 第113-117页 |
| 5.3 Application Examples | 第117-139页 |
| 5.3.1 Parametric Modeling of a Microwave Junction | 第117-124页 |
| 5.3.2 Parametric Modeling of a Diplexer | 第124-130页 |
| 5.3.3 Parametric Modeling of a Microwave Cavity Filter | 第130-139页 |
| 5.4 Conclusion | 第139-140页 |
| Chapter 6 Conclusions and Future Research | 第140-144页 |
| 6.1 Conclusions | 第140-141页 |
| 6.2 Future Research | 第141-144页 |
| References | 第144-156页 |
| List of Publications | 第156-159页 |
| Acknowledgements | 第159-160页 |
