【正文】
ffice 開發(fā)的應(yīng)用程序中,為圖形編程提供繪圖界面,并允許開發(fā)人員控制集成的 Visio 用戶界面。本文的依賴矩陣顯示的數(shù)據(jù)表格,和指標(biāo)分配模塊的線性插值法加權(quán)參數(shù)設(shè)置表格使用了該組件。3.CGridCtrl 表格數(shù)據(jù)顯示組件CGridCtrl 類是 網(wǎng)站的創(chuàng)始人之一 Chris Maunder 開發(fā)的一個(gè)數(shù)據(jù)表格顯示編輯開源控件,該控件繼承于 MFC 的 CWnd 類,其功能全面,簡單易用。除此,其數(shù)據(jù)結(jié)構(gòu)明晰易懂,代碼簡潔明了易用。 Greiner 聯(lián)合開發(fā)的純數(shù)據(jù)樹形數(shù)據(jù)結(jié)構(gòu)。本軟件將可測性模型數(shù)據(jù)存儲(chǔ)為 XML(Extensible Markup Language,即可擴(kuò)展標(biāo)記語言)格式的文件,并通過 CMarkup 類來實(shí)現(xiàn)工程文件數(shù)據(jù)的讀寫操作。用它來創(chuàng)建新的 XML 文檔,分析修改現(xiàn)存的 XML 文檔,非常方便易用。1.CMarkup XML 文件讀寫組件CMarkup 類是由 First Objective Software Inc 用 C++語言開發(fā)的非常經(jīng)典的XML 文件讀寫工具。除此,它還提供了詳細(xì)的程序開發(fā)手冊,介紹其定義類庫和函數(shù)的接口,使應(yīng)用程序開發(fā)變得簡單。MFC 中封裝了大部分Windows API(Application Programming Interface,應(yīng)用程序編程接口)函數(shù)和Windows 控件,它包含的功能涉及到整個(gè) Windows 操作系統(tǒng)。它支持過程化程序設(shè)計(jì)、數(shù)據(jù)抽象、面向?qū)ο蟪绦蛟O(shè)計(jì)、泛型程序設(shè)計(jì)等多種程序設(shè)計(jì)風(fēng)格 [21]。 軟件平臺(tái)介紹本軟件是在 VC++ 編程平臺(tái)下開發(fā),并且以 Windows 操作系統(tǒng)為運(yùn)行平臺(tái)。第二章 軟件結(jié)構(gòu)總體設(shè)計(jì)5第二章 軟件結(jié)構(gòu)總體設(shè)計(jì)本章從軟件整體結(jié)構(gòu)設(shè)計(jì)的角度出發(fā),先介紹本軟件的編程開發(fā)平臺(tái)和編程過程中引用的第三方開源組件,及本軟件具備的功能。第六章,軟件測試與驗(yàn)證本章以“反坦克導(dǎo)彈發(fā)動(dòng)機(jī)系統(tǒng)”測試驗(yàn)證實(shí)例,建立本軟件的可測性模型,對本軟件的各功能模塊和整體性能進(jìn)行一一測試。對可測性指標(biāo)分配做了簡要介紹,還介紹了經(jīng)驗(yàn)法、線性插值法兩種分配方法。重點(diǎn)闡述了最優(yōu)故障診斷樹生成模塊的軟件設(shè)計(jì)和編程實(shí)現(xiàn),包括模塊的組成結(jié)構(gòu)設(shè)計(jì),霍夫曼樹的構(gòu)造、霍夫曼平均碼長的求取實(shí)現(xiàn),AO*算法的編程實(shí)現(xiàn),診斷樹結(jié)果的圖形化交互顯示界面的設(shè)計(jì)和實(shí)現(xiàn)。該方法在依賴矩陣基礎(chǔ)上,只需采用現(xiàn)有測試方法就能評估驗(yàn)證虛警率,希望能為虛警率評估提供一種新的思路,并編程實(shí)現(xiàn)了該方法。第三章,可測性指標(biāo)分析模塊設(shè)計(jì)與實(shí)現(xiàn)本章介紹了常見可測性指標(biāo)(故障檢測率、故障隔離率、未檢測故障集、故障模糊集、隱藏故障、掩蓋故障、冗余測試) 、動(dòng)態(tài)性能指標(biāo)(加權(quán)故障檢測率、加權(quán)故障隔離率、平均模糊度)的定義及計(jì)算方法,并編程實(shí)現(xiàn)了這些指標(biāo)的計(jì)算方法和顯示可測性指標(biāo)結(jié)果的報(bào)告輸出。還介紹了軟件具備的功能,并將其劃分為系統(tǒng)級可視化建模、可測性指標(biāo)分析、最優(yōu)故障診斷樹生成、可測性指標(biāo)分配四個(gè)大功能模塊。軟件總體設(shè)計(jì)與集成指標(biāo)分析 模塊設(shè)計(jì)與實(shí)現(xiàn)故障診斷樹 模塊設(shè)計(jì)與實(shí)現(xiàn)指標(biāo)分配 模塊設(shè)計(jì)與實(shí)現(xiàn)系統(tǒng)級可視化建模模塊圖 11 本文研究內(nèi)容根據(jù)本文研究內(nèi)容及可測性軟件的關(guān)聯(lián)性,本文的結(jié)構(gòu)安排如下: 第一章,緒論介紹可測性技術(shù)的定義及課題背景和國內(nèi)外可測性軟件概況。 本文研究內(nèi)容及結(jié)構(gòu)安排本論文的主要研究工作是可測性分析與輔助設(shè)計(jì)軟件的總體結(jié)構(gòu)設(shè)計(jì)與實(shí)現(xiàn),和可測性指標(biāo)分析模塊、最優(yōu)故障診斷樹模塊、可測性指標(biāo)分配模塊的設(shè)計(jì)與實(shí)現(xiàn)。因?yàn)榭蓽y性設(shè)計(jì)是軍用電子設(shè)備研制必要指標(biāo)之一,而且未考慮可測性設(shè)計(jì)的電子設(shè)備大大增加了研制單位的維護(hù)第一章 緒論3費(fèi)用,所以這些單位迫切希望能對現(xiàn)有電子系統(tǒng)和設(shè)備進(jìn)行深入的可測性分析,為設(shè)備的后期測試、維護(hù)提高故障診斷依據(jù)。從技術(shù)層面上看,主要是缺乏系統(tǒng)的可測性設(shè)計(jì)指南、理論指導(dǎo),和缺乏有效的關(guān)于可測性設(shè)計(jì)的計(jì)算機(jī)輔助設(shè)計(jì)與仿真軟件工具 [1120]。1990 年 4 月發(fā)布的航標(biāo) HB643790《電子系統(tǒng)和設(shè)備的可測試性大綱 》 ,1995 年 10 月發(fā)布國軍標(biāo) GJB254795《裝備測試性大綱》 。除此,TEAMS 在可測性輔助設(shè)計(jì)方面功能不夠完善,而 eXpress 的學(xué)科關(guān)聯(lián)也太多,缺乏對系統(tǒng)可測性指標(biāo)分析和輔助設(shè)計(jì)的針對性。eXpress 強(qiáng)調(diào)診斷性設(shè)計(jì),其中的診斷信息可產(chǎn)生多個(gè)診斷報(bào)告,這些報(bào)告包括檢測覆蓋率報(bào)告,詳細(xì)故障組統(tǒng)計(jì),功能概率,故障模式概率等。產(chǎn)品的可測性經(jīng)過 TEAMS 分析后,生成文本和圖形的可測性報(bào)告。TEAMS 是基于模型的分析工具,這些模型涵蓋了系統(tǒng)的組成結(jié)構(gòu)、互聯(lián)、測試布局、故障模式等信息。TEAMS 主要用于復(fù)雜系統(tǒng)的測試時(shí)序生成和可測性設(shè)計(jì)、分析。隨著計(jì)算機(jī)軟件工程的發(fā)展、武器系統(tǒng)測試和診斷要求提高,目前的可測性輔助設(shè)計(jì)和評價(jià)軟件工具不僅可以用于系統(tǒng)的測試和診斷設(shè)計(jì),而且可以生成相應(yīng)的測試序列和故障診斷引擎,以提高系統(tǒng)的維護(hù)性、可靠性。這些工具包括:ASTEP 、CAFIT、DTA 、INATE、LOGMOD 、SCOAP、STAMP 、ACE、WATA 等。考慮可測性的綜合診斷技術(shù)已應(yīng)用在正在研制的新一代武器系統(tǒng)中,如軍用運(yùn)輸機(jī) C1戰(zhàn)斗機(jī) F2轟炸機(jī) B2。為與綜合診斷的協(xié)調(diào),美國國防部于 1993 年 2 月頒發(fā) MIL STD2165A《系統(tǒng)和設(shè)備的可測性大綱》 ,大綱將可測試性作為與可靠性及維修性等同的設(shè)計(jì)要求,并規(guī)定了可測試性分析、設(shè)計(jì)及驗(yàn)證的要求及實(shí)施方法??傊?,可測性設(shè)計(jì)是與產(chǎn)品功能設(shè)計(jì)并行的復(fù)雜工作,不能僅從設(shè)計(jì)階段的投入衡量,還要從它在產(chǎn)品整個(gè)壽命周期獲得的利益考慮。具體是指在系統(tǒng)、分系統(tǒng)、設(shè)備、組件和部件的設(shè)計(jì)過程中,通過綜合考慮并實(shí)現(xiàn)測試的可控性與可觀測性、初始化與可達(dá)性以及和外部測試設(shè)備兼容性等,達(dá)到可測性要求的設(shè)計(jì)過程。以提高可測性為目的進(jìn)行的設(shè)計(jì)被稱為可測性設(shè)計(jì)(DFT: Design For Testability)。在這種情況下,人們提出了可測性設(shè)計(jì)概念,在設(shè)計(jì)階段就考慮產(chǎn)品的后期測試維護(hù)問題,把降低測試維護(hù)代價(jià)的要求納入設(shè)計(jì)規(guī)范,并通過可測性指標(biāo)來檢驗(yàn)和改進(jìn)設(shè)計(jì)。關(guān)鍵詞:可測性,指標(biāo)分析,故障診斷樹,指標(biāo)分配,虛警率ABSTRACTIIIABSTRACTTestability is a key factror for a product design through which the product’s status (work, unwork, performance degradation) can be determined timely and accurately, and the internal fault can be isolated. Design for testability (DFT) has bee an important part in the design of weapons and equipments because DFT plays a decisive role in imroving supportability and reducing test costs for weapons and equipments. Although China promulgated the Outline of test equipment in 1995, the DFT status applied in current electronic equipment lags far behind testability requirements for the new generation equipments due to a lack of an effective puteraided tool for DFT design and evaluation, Therefore, in support by some projects of the State, based on the TEAMS—a mercial software for testability, we do a thorough research in software architecture design and key modules design for DFT software. Our DFT software has been verified in some related units. Main contents in this thesis are as follows: 1. The architecture of the software is designed. The software is divided into four functional modules . systemlevel visual modeling, testability index analysis, optimal fault diagnosis tree generation, testability indicator distribution. Using VC++ as the development platform, we designed the architecture of the software. In addition, we designed a multinode model tree and some dependencematrix data structures that play a key role when data transmission between modules. 2. A false alarm rate evaluation approach based on multitest information from a dependencematrix is proposed. Based on dependencematrix, this method can verify the false alarm rate only use existing test methods. Thus this method provides a new way for evaluating false alarm rate. In addition, the calculation methods for mon testability index (fault detection rate, fault isolation rate, undetected fault sets, fault ambiguity sets, hidden fault sets, masking fault sets, redundant test sets) and for dynamic testability index (weighted fault detection rate, the weighted fault isolation rate, the mean ambiguity degree) are implemented and optimized. 3. Based on AO* heuristic search algorithm, the optimal method for fault diagnosis tree generation is implemented. AO* heuristic search algorithm is the basis of testability ABSTRACTIVanalysis and design optimization, but its algorithm description is plexity and it has plexity putation. In order to reduce resource overhead when the program runs, the program identify all the fault sets and usable the tests by the sequence of ID numbers on the time dependencematrix. This method avoids redundant information in processing meaningless during the operation. Moreover, the method reduces the unnecessary requirement for memory and CPU overhead to some extent. In addition, we also design and implement an interactive graphical display interface for the results of the diagnostic tree.4. Because of a shortage of index distribution in TEAMS software, the paper proposes to integrate a testability index distribution module into our testability software. First, an index distribution is briefly r