本文選題：印刷電路板　切入點：故障診斷　出處：《上海工程技術(shù)大學(xué)》2015年碩士論文

【摘要】：隨著科學(xué)技術(shù)的發(fā)展,印刷電路板(Printed Circuit Board,PCB)需求量的日益增加,板載電子電路系統(tǒng)也變得越來越復(fù)雜,對出廠前故障診斷的要求也越來越高,如何快速、高效地檢測PCB板的質(zhì)量,已經(jīng)成為生產(chǎn)企業(yè)不得不高度重視的問題。然而,傳統(tǒng)的人工檢測方法存在著勞動強(qiáng)度大、可靠性差(易發(fā)生漏檢和誤檢)等缺點,直觀、高效的智能檢測方法逐步替代人工檢測方法成為必然趨勢。本文對PCB的智能診斷和可視化技術(shù)進(jìn)行研究,首先介紹了PCB故障診斷技術(shù)的背景及其研究意義,比較全面地闡述了PCB的故障診斷技術(shù),包括PCB故障的分類、診斷PCB故障的常用方法等,著重闡述了本文所采用的PCB功能測試法;并針對傳統(tǒng)的PCB故障診斷技術(shù)過程繁瑣、不直觀等缺點,提出了基于工控軟件Lab VIEW(Laboratory Virtual Instrument Engineering Workbench)和PLC(Programmable Logic Controller,PLC)的PCB故障診斷可視化系統(tǒng)。該系統(tǒng)將采用Lab VIEW和PLC作為主要的開發(fā)平臺,為了解決PCB故障元件的精確定位問題,本文利用Visual C++6.0封裝Visio 2007 Active X Control繪圖控件,繼而將Active X控件嵌入到Lab VIEW中,通過Lab VIEW友好的人機(jī)交互界面定位被檢測到的故障元件,以實現(xiàn)系統(tǒng)的可視化;為更好地建立PCB故障智能診斷與可視化系統(tǒng)的模型,本文還引入了隱馬爾科夫模型(Hidden Markov Models,HMM),并且對HMM要解決的三個基本問題,即評估問題、解碼問題和訓(xùn)練問題,以及針對這三個問題的Foreword-Backward算法、Viterbi算法和Baum-Welch算法進(jìn)行了詳述;為解決HMM初始參數(shù)的確立及訓(xùn)練的難題,引入了遺傳算法(Genetic Algorithm,GA)和時域上相關(guān)的隱馬爾科夫模型(Time Correlation HMM,TC-HMM);隨后,分別介紹了遺傳編碼、適應(yīng)度函數(shù)、遺傳算子以及時域上相關(guān)的隱馬爾科夫模型算法的設(shè)計;為了能更加清晰地描述PCB故障智能診斷與可視化系統(tǒng)的構(gòu)建,本文重點闡述了基于遺傳算法與TC-HMM相結(jié)合的故障診斷方法的實現(xiàn),最終利用遺傳算法改進(jìn)和優(yōu)化時域上相關(guān)的隱馬爾科夫模型的參數(shù)初始化訓(xùn)練過程,以建立本文所需的PCB故障智能診斷模型。最后,為驗證本文所提出的算法和所建立模型的綜合性能,建立了智能診斷模型的實驗平臺,通過大量實驗和測試驗證診斷系統(tǒng)的性能。實際結(jié)果表明該診斷系統(tǒng)達(dá)到了預(yù)期設(shè)計目的,提高了故障診斷效率,并實現(xiàn)了故障定位的可視化,具有自動化程度高、操作簡單、直觀等特點,適用于大批量PCB板的故障檢測。
[Abstract]:With the development of science and technology, the demand for printed Circuit boards is increasing day by day, and the on-board electronic circuit system is becoming more and more complex.Has become the production enterprise has to attach great importance to the problem.However, the traditional manual detection method has many disadvantages, such as high labor intensity, poor reliability (easy to miss detection and false detection) and so on. Therefore, it is inevitable to replace manual detection method with intuitive and efficient intelligent detection method.In this paper, the intelligent diagnosis and visualization technology of PCB is studied. Firstly, the background and significance of PCB fault diagnosis technology are introduced, and the fault diagnosis technology of PCB, including the classification of PCB fault, is expounded comprehensively.In this paper, the PCB function test method used in this paper is expounded, and the disadvantages of the traditional PCB fault diagnosis technology are discussed, such as tedious, not intuitive, etc.This paper presents a visual PCB fault diagnosis system based on Lab VIEW(Laboratory Virtual Instrument Engineering Workbench and PLC(Programmable Logic Controller.This system will adopt Lab VIEW and PLC as the main development platform. In order to solve the problem of accurate location of PCB fault components, this paper encapsulates the Visio 2007 Active X Control drawing control by Visual C 6.0, and then embeds Active X control into Lab VIEW.In order to realize the visualization of the system by locating the detected fault elements through the friendly man-machine interface of Lab VIEW, the model of the intelligent diagnosis and visualization system for PCB fault is established.In this paper, Hidden Markov models are also introduced, and three basic problems to be solved by HMM, namely, evaluation problem, decoding problem and training problem, as well as the Foreword-Backward algorithm and Baum-Welch algorithm for these three problems are described in detail.In order to solve the problem of establishing and training the initial parameters of HMM, the genetic algorithm (GA) and the time Correlation HMMC-HMMM model in time domain are introduced, and the genetic coding and fitness function are introduced respectively.The design of genetic operator and hidden Markov model algorithm in time domain, in order to describe the construction of intelligent diagnosis and visualization system of PCB fault more clearly,This paper focuses on the implementation of fault diagnosis method based on genetic algorithm and TC-HMM. Finally, genetic algorithm is used to improve and optimize the parameter initialization training process of the related hidden Markov model in time domain.In order to establish the PCB fault intelligent diagnosis model needed in this paper.Finally, in order to verify the proposed algorithm and the comprehensive performance of the established model, an experimental platform of intelligent diagnosis model is established, and the performance of the diagnostic system is verified by a large number of experiments and tests.The practical results show that the system achieves the desired design purpose, improves the efficiency of fault diagnosis, and realizes the visualization of fault location. It has the characteristics of high automation, simple operation, intuitive operation, etc.It is suitable for the fault detection of mass PCB board.
【學(xué)位授予單位】：上海工程技術(shù)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：TN41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 劉思久,汪秀豐,董景;基于虛擬儀器方式的PCB功能測試[J];電測與儀表;2004年01期

2 張琪;侯加林;閆銀發(fā);李天華;;基于虛擬儀器的電路板故障檢測與診斷系統(tǒng)的研究[J];電子測量與儀器學(xué)報;2011年02期

3 張光軼,許愛強(qiáng),宋振宇;三種測試總線的總線控制方法研究[J];儀器儀表用戶;2005年03期

4 王吉林;曾曉宏;;高密度印刷線路板功能測試思路探析[J];高校實驗室工作研究;2007年01期

5 華夏;;Labview與PLC在工業(yè)場合結(jié)合應(yīng)用探索[J];科技創(chuàng)新與應(yīng)用;2013年06期

6 王恒;吳媛媛;;集成電路在線功能測試技術(shù)研究[J];艦船電子工程;2007年03期

7 王躍鋼;紅外熱成像診斷電路板的動態(tài)激勵方法[J];激光與紅外;2002年01期

8 郝杰,李星;對經(jīng)典隱馬爾可夫模型的經(jīng)驗性改進(jìn)[J];計算機(jī)工程與應(yīng)用;2001年11期

9 房曉杰;CPLD及邊緣掃描測試技術(shù)在電動車組輔助系統(tǒng)中的應(yīng)用研究[J];計算機(jī)應(yīng)用與軟件;2003年03期

10 王振鵬,段晨東,何正嘉;基于虛擬儀器技術(shù)實現(xiàn)PLC控制系統(tǒng)的狀態(tài)監(jiān)測和故障診斷[J];計算機(jī)測量與控制;2004年08期

相關(guān)會議論文 前1條

1 楊宣訪;夏立;;電子裝備路內(nèi)測試診斷技術(shù)研究[A];艦船電子裝備維修理論與應(yīng)用——中國造船工程學(xué)會電子修理學(xué)組第四屆年會暨信息裝備保障研討會論文集[C];2005年

相關(guān)碩士學(xué)位論文 前10條

1 錢晨;隱馬爾科夫模型演化下的隱組檢測[D];西安電子科技大學(xué);2009年

2 商曄;隱馬爾可夫模型參數(shù)訓(xùn)練的改進(jìn)及在股市預(yù)測中的應(yīng)用[D];上海交通大學(xué);2011年

3 陳希;網(wǎng)絡(luò)化邊緣掃描測試控制器的研制與系統(tǒng)組建[D];電子科技大學(xué);2006年

4 常征;功能測試中自動化測試框架的分析與應(yīng)用[D];北京林業(yè)大學(xué);2007年

5 張春霞;基于隱馬爾科夫模型的細(xì)胞分裂檢測[D];中國海洋大學(xué);2008年

6 黃建喜;基于虛擬儀器的電路板自動測試系統(tǒng)的設(shè)計與實現(xiàn)[D];華中科技大學(xué);2008年

7 保坤;電路板故障診斷系統(tǒng)可視化開發(fā)與運行環(huán)境的設(shè)計與實現(xiàn)[D];電子科技大學(xué);2010年

8 游望星;基于DSP的電路故障診斷系統(tǒng)研究與開發(fā)[D];湖南大學(xué);2010年

9 楊筆鋒;基于改進(jìn)訓(xùn)練算法的HMM語音識別技術(shù)研究[D];湖南大學(xué);2010年

10 王猛;基于LabVIEW的電路板故障測試系統(tǒng)的設(shè)計[D];北京交通大學(xué);2012年

，

本文編號：1693228