【正文】 and IWPT is effective for feature extraction of hard fault diagnosis.That support vector machine (SVM) is originally designed for twoclass classification, but pattern classification question belongs to multiclassification. Aiming at shorts of several mon multiclassification methods based on SVMs: 1versusrest (1Vr), 1versus1 (1V東北大學(xué)碩士學(xué)位論文 Abstract－VII －1) and decision directed acyclic graph (DDAG), four binarytree support vector machines (BTSVMs)multiclassification methods are proposed in this paper: slantwise binary tree support vector machines (SBTSVMs), Random binary tree support vector machines (RBTSVMs), pleted binary tree support vector machines (CBTSVMs) and adaptive binary tree support vector machines (ABTSVMs). Simulation results show that BTSVMs multiclassifiers are feasible to solve fault diagnosis problem, and obtain high classification precision and speed.With MATLAB platform, fault diagnosis system of analog circuits is developed. Several modules of fault diagnosis system are designed and the improved BTSVMs algorithms are realized.Keywords: Analog Circuit。 Fault Diagnosis。 Wavelet Packet。 Support Vector Machine。 Binary Tree。 Multiclassification 東北大學(xué)碩士學(xué)位論文 目 錄－VIII－目 錄獨(dú)創(chuàng)性聲明 .................................................................................................................................I摘 要 ..........................................................................................................................................IIABSTRACT .............................................................................................................................IV第一章 緒 論 .............................................................................................................................1 課題研究背景及意義..........................................................................................................1 模擬電路故障診斷技術(shù)的發(fā)展及研究現(xiàn)狀......................................................................2 模擬電路故障診斷方法分類..............................................................................................3 課題研究所做的主要工作..................................................................................................5第二章 模擬電路故障診斷 .......................................................................................................7 模擬電路故障診斷的基本概念..........................................................................................7 常用的模擬電路故障診斷方法..........................................................................................8 測前模擬方法...........................................................................................................8 測后模擬法...............................................................................................................8 逼近法和人工智能...................................................................................................9 模擬電路故障診斷的模式識別過程................................................................................10 本章小結(jié)............................................................................................................................12第三章 電路仿真工具— PSPICE 的分析及應(yīng)用 ..................................................................13 PSpice 的發(fā)展 ....................................................................................................................13 PSpice 的內(nèi)容 ....................................................................................................................14 PSpice 的功能介紹和應(yīng)用分析 .......................................................................................15 PSpice 的編程技術(shù) ................................................................................................15 PSpice 的特殊分析方法 ........................................................................................16 本章小結(jié)............................................................................................................................18第四章 小波包分析理論及基于小波包分析的信號特征提取方法 .....................................19 小波變換的優(yōu)點(diǎn)................................................................................................................19 一維連續(xù)小波變換理論....................................................................................................20 小波函數(shù)的選擇................................................................................................................21 小波包理論........................................................................................................................23 小波包定義.............................................................................................................23 小波包性質(zhì).............................................................................................................24東北大學(xué)碩士學(xué)位論文 目 錄－IX － 小波包的空間分解.................................................................................................25 小波包算法.............................................................................................................26 基于小波包變換的信號特征提取方法............................................................................26 最優(yōu)小波包特征提取方法.....................................................................................28 不完全小波包特征提取方法.................................................................................29 本章小結(jié)............................................................................................................................30第五章 基于神經(jīng)網(wǎng)絡(luò)和支持向量機(jī)的多類分類方法 .........................................................31 BP 神經(jīng)網(wǎng)絡(luò)的基本算法 ...................................................................................................31 統(tǒng)計(jì)學(xué)習(xí)理論....................................................................................................................33 支持向量機(jī)原理................................................................................................................33 支持向量機(jī)優(yōu)化算法...............................................................