神經(jīng)網(wǎng)絡(luò)專家系統(tǒng)在凝汽器故障診斷中的應(yīng)用研究.pdf

1、華北電力大學(xué)碩士學(xué)位論文摘要凝汽器運行的熱力性能對汽輪機機組運行的安全性與經(jīng)濟性有很重要的影響，研究凝汽器故障診斷方法，實施故障診斷技術(shù)對提高譏組經(jīng)濟運行水平有著重要的理論意義和工程實用價值。本文論述了火電廠凝汽設(shè)備的運行特性，分析了凝汽系統(tǒng)的常見故障，重點分析了凝汽器真空惡化的原因及征兆全面闡述了神經(jīng)網(wǎng)絡(luò)和專家系統(tǒng)的基本理論和系統(tǒng)結(jié)構(gòu)，設(shè)計了神經(jīng)網(wǎng)絡(luò)和專家系統(tǒng)結(jié)合在一起的一種新方法，使二者更好地應(yīng)用于故障診斷領(lǐng)域。以BP神經(jīng)網(wǎng)絡(luò)為基礎(chǔ)

2、建立凝汽器故障診斷專家系統(tǒng)，對網(wǎng)絡(luò)進行訓(xùn)練，然后利用所學(xué)成的網(wǎng)絡(luò)，對凝汽器故障數(shù)據(jù)進行處理，得到了凝汽器具體的故障情況。本文利用MATLAB作為開發(fā)工具，利用神經(jīng)網(wǎng)絡(luò)的學(xué)習(xí)規(guī)則，采用自適應(yīng)學(xué)習(xí)率BP算法完成故障知識庫的組建使用Delphi7.O開友了基于BP網(wǎng)絡(luò)的汽輪機組凝汽系統(tǒng)故障診斷系統(tǒng)應(yīng)用軟件。關(guān)鍵詞:凝汽器故障診斷神經(jīng)網(wǎng)絡(luò)專家系統(tǒng)ABSTRACTThethennodynamiecharaeteristieofcondensing

3、systemhasaheavyimPactiononthesecurityandeeonomyofarunningturbine.It15imPertantthatestablishingthemonitoringandoPtimizationmodelffaultdiagnosisandimProvirgtheoPerationlevel.InthisPapef，eondensingsystem，5Performance15discu

4、ssed，andeondensingsystem’seommonfaultsareanalyZedandtheeauseandsignoofeondenser，5vacuu們nfallingareernPhasized.ThebasictheoryandsystemstnjctureofZ、rtifieialNeuralNeti刀orkandExPertSystemareinirodueedinthisPaper.Itdesignsan

5、ewInethodwhieheombinestheExPertsSystemwiththeArtifieialNeuralNetworktomakeapPli，ationffaultdiagnosisfieldbetter.Combinedwitheaehsituationindetail，condenserdia乏筍0515ExPertSystembasedonBPNeuralNetwk15construeted.TrainsNetw

6、k，thendealswithdiagnosisdatumthroughmakinguseoftheleamedNetwork，50itobtainsdiagnosissituationindctail.WithMATLAB，makinguseoftheNeuralNetw(，rkleamingrules，the知owledgeeonstruction15eomPleted:WithDELPHI7.0，幾ultdia毛nosisaPPl

7、ieationsoftwareftheeondensingsystemwasdeveloPed.ShanXusheng(ThermalP.)werengineering)DireetedbyProf.ZhangGuangKeywords:eondenserfaultdiagnosisNeuralNetworkExPertSystem華北電力大學(xué)碩士學(xué)位論文3.2.2專家系統(tǒng)的開發(fā)步驟...........................

8、........……283.2.3專家系統(tǒng)的知識表示....................................……293.2.4專家系統(tǒng)的知識獲取....................................……293.3神經(jīng)網(wǎng)絡(luò)和專家系統(tǒng)的融合...................................……303.3.1神經(jīng)網(wǎng)絡(luò)與專家系統(tǒng)的差異.............................

9、.……303.3.2專家系統(tǒng)與神經(jīng)網(wǎng)絡(luò)的互補.............................……313.3.3專家系統(tǒng)和神經(jīng)網(wǎng)絡(luò)的融合方式..........................……313.4本章小結(jié)...................................................……32第四章凝汽器故障診斷軟件系統(tǒng)的實現(xiàn)..............................……334.1

10、BP神經(jīng)網(wǎng)絡(luò)學(xué)習(xí)算法的選定..................................……334.2BP神經(jīng)網(wǎng)絡(luò)構(gòu)造............................................……354.3凝汽器故障特征信號的選取...................................……364.4Matlab神經(jīng)網(wǎng)絡(luò)工具箱......................................……

11、394.5BP神經(jīng)網(wǎng)絡(luò)的訓(xùn)練..........................................……394.6故障診斷步驟...............................................……444.7診斷實例...................................................……444.8凝汽系統(tǒng)運行工況監(jiān)視及故障診斷系統(tǒng)..................

12、.......……454.8.1系統(tǒng)的物理結(jié)構(gòu).......................................……454.8，2機組運行工況監(jiān)視系統(tǒng)界面............................……454.8.3凝汽系統(tǒng)故障診斷系統(tǒng)界面.............................……484.9本章小結(jié)................................................

13、...……50第五章結(jié)論與展望................................................……51參考文獻..........................................................……52致謝...........................................................……54在學(xué)習(xí)期間發(fā)表的學(xué)術(shù)論文和參加科研情況...