【正文】 ............................8第 3 章 小波分析及信號(hào)處理 .................................................................................................9 小波分析在信號(hào)處理中的應(yīng)用現(xiàn)狀 ..............................................................................9 小波分析理論 ................................................................................................................10 小波分析的基本概念及特點(diǎn) ...................................................................................10 多分辨率分析 ..........................................................................................................113．3 小波包分析 .................................................................................................................12 小波包的定義 ..........................................................................................................12 小波包的子空間分解 ..............................................................................................12 小波包的分解與重構(gòu)算法 ......................................................................................12 小波基函數(shù)的選擇 ........................................................................................................13 信號(hào)的小波包降噪 ........................................................................................................15 小波包降噪的方法和步驟 ......................................................................................15 降噪效果的評(píng)價(jià)標(biāo)準(zhǔn) ..............................................................................................16v 故障信號(hào)的小波包降噪 ..........................................................................................17 故障信號(hào)的特征提取 ....................................................................................................19第 4 章 BP 神經(jīng)網(wǎng)絡(luò)及在故障診斷中的應(yīng)用 .....................................................................21 人工神經(jīng)網(wǎng)絡(luò)在故障診斷中的應(yīng)用現(xiàn)狀 ....................................................................21 BP 神經(jīng)網(wǎng)絡(luò) ...................................................................................................................21 BP 神經(jīng)網(wǎng)絡(luò)的結(jié)構(gòu) .................................................................................................21 BP 算法描述 .............................................................................................................22 BP 神經(jīng)網(wǎng)絡(luò)的優(yōu)缺點(diǎn) .............................................................................................24 BP 神經(jīng)網(wǎng)絡(luò)的設(shè)計(jì) .......................................................................................................24 學(xué)習(xí)樣本的確定 ......................................................................................................24 網(wǎng)絡(luò)層數(shù)的考慮 ......................................................................................................25 輸入、輸出和隱含層的設(shè)計(jì) ...................................................................................25 激勵(lì)函數(shù)的選擇 ......................................................................................................26 網(wǎng)絡(luò)初始值的選取 ...................................................................................................26 學(xué)習(xí)算法的選擇 .......................................................................................................26 學(xué)習(xí)速率 ..................................................................................................................26 訓(xùn)練停止條件 ..........................................................................................................28 BP 神經(jīng)網(wǎng)絡(luò)在故障診斷中的應(yīng)用 ...............................................................................28 BP 神經(jīng)網(wǎng)絡(luò)故障診斷結(jié)果分析 ...................................................................................30第 5 章 GA+BP 神經(jīng)網(wǎng)絡(luò)及其在故障診斷中的應(yīng)用 .........................................................31 遺傳算法簡(jiǎn)介 ................................................................................................................31 標(biāo)準(zhǔn)遺傳算法 ................................................................................................................31 遺傳算法的特點(diǎn) ............................................................................................................32 GA+BP 算法設(shè)計(jì) ...........................................................................................................33 遺傳編碼與解碼 ......................................................................................................33 適應(yīng)度函數(shù)的設(shè)計(jì) ..................................................................................................36 遺傳算子的設(shè)計(jì) ......................................................................................................36 控制參數(shù)的選擇 ......................................................................................................39vi GA+BP 算法的步驟 .................................................................................................41 GA+BP 神經(jīng)網(wǎng)絡(luò)在故障診斷中的應(yīng)用 .......................................................................42 網(wǎng)絡(luò)的設(shè)計(jì) ...............................................................................................................42 故障診斷仿真實(shí)驗(yàn) ..................................................................................................43 故障診斷結(jié)果分析 ..................................................................................................45 BP 神經(jīng)網(wǎng)絡(luò)與 GA+BP 神經(jīng)網(wǎng)絡(luò)故障診斷結(jié)果對(duì)比分析 ........................................46結(jié)論 .........................................................................................................................................47參考文獻(xiàn) .................................................................................................................................49攻讀學(xué)位期間取得的成果 .....................................................................................................51致 謝 .......................................................................................................................................51中國(guó)石油大學(xué)（華東）碩士學(xué)位論文1第 1 章 緒論 故障診斷的意義隨著科技日新月異的發(fā)展，現(xiàn)代化生產(chǎn)所用的生產(chǎn)裝置也趨向大型化、高速高效化、自動(dòng)化和連續(xù)化，人們對(duì)設(shè)備的要求不僅是性能好，效率高，還要求在運(yùn)行過(guò)程中無(wú)故障或少故障。GA+BP神經(jīng)網(wǎng)絡(luò)與BP神經(jīng)網(wǎng)絡(luò)故障診斷的結(jié)果對(duì)比后可以看到：第一，