【正文】
LIM(軟件方程式模型) SLIM 方法是在 20 世紀(jì) 70 年代后期由 QSM 組織的 Putnam 開發(fā)的,它是一個(gè)動(dòng)態(tài)的多變量模型。 COCOMO 方法又根據(jù)待開發(fā)軟件的特點(diǎn),分為組織式、半分離式和嵌入式三種模式。參數(shù)估計(jì)方法的準(zhǔn)確度可以通過校正因子處理而得到提高。由于軟件進(jìn)度估計(jì)總是依賴于軟件工作量估計(jì)和可以投入的軟件人力資源,在人力資源投入策略確定后,軟件開發(fā)工作量與軟件項(xiàng)目進(jìn)度的對應(yīng)關(guān)系就確定了。在項(xiàng)目可以開始前,管理者和軟件小組必須估計(jì)將要完成的工作、所需要的資源以及從開始到完成所需要的時(shí)間。 不難發(fā)現(xiàn),軟件項(xiàng)目風(fēng)險(xiǎn)管理的研究與項(xiàng)目進(jìn)度計(jì)劃的研究是有交集的,在考慮項(xiàng)目風(fēng)險(xiǎn)時(shí),進(jìn)度風(fēng)險(xiǎn)通常是考慮 的重點(diǎn),在制定項(xiàng)目進(jìn)度計(jì)劃時(shí),要考慮達(dá)到進(jìn)度目標(biāo)可能遇到的風(fēng)險(xiǎn)。 Fairley( 1994)在他的《軟件項(xiàng)目的風(fēng)險(xiǎn)管理》[8]一文中驗(yàn)證了 Boehm 的方法在電信軟件項(xiàng)目中的應(yīng)用,他充分利用了 COCOMO 成本估算模型來估計(jì)風(fēng)險(xiǎn)因子對預(yù)算的影響,并且證明了人們可以利用統(tǒng)計(jì)學(xué)方法求出可能產(chǎn)生結(jié)果的預(yù)期范圍。 在軟件領(lǐng)域,關(guān)于項(xiàng)目風(fēng)險(xiǎn)管理和項(xiàng)目進(jìn)度計(jì)劃主題的文獻(xiàn)著作很多。當(dāng)不切實(shí)際的最終期限臨近時(shí) ,為了盡快完成項(xiàng)目,全體開發(fā)人員可能要忍受被施加的附加壓力。 產(chǎn)品質(zhì)量影響。 Kemerer( 1991) [8]認(rèn)為進(jìn)度壓力常常在項(xiàng)目的后期出現(xiàn),并對項(xiàng)目帶來三個(gè)主要方面的影響: 1失控項(xiàng)目的定義 [KPMG 1995]:軟件失控項(xiàng)目是顯著未能實(shí)現(xiàn)目標(biāo)和(或)至少超出原定預(yù)算 30%的項(xiàng)目。 Jones 認(rèn)為軟件項(xiàng)目計(jì)劃不合理是軟件項(xiàng)目交付延期的主要原因。 造成以上現(xiàn)象的原因有很多, Jones( 1994) [23]針對交付延期和預(yù)算超支的現(xiàn)象,歸納出以下四個(gè)根本原因: 在項(xiàng)目初始估計(jì)時(shí),進(jìn)度 /成本就是不可能達(dá)到的目標(biāo),但項(xiàng)目還是如期啟動(dòng)了; 在項(xiàng)目進(jìn)度 /成本確定后,項(xiàng)目范圍發(fā)生了變化; 項(xiàng)目估計(jì)和計(jì)劃的方法不合理; 企業(yè)沒有收集有用的歷史數(shù)據(jù)。s needs. The results indicate that the simulation model can help project manager to prepare his risk management plan and schedule plan effectively and efficiently. Key words: risk factor, simulation model, risk management plan, schedule plan 南京大學(xué)碩士論文 4 目錄 第一章 緒論 .................................................................................................................... 1 本文研究的背景及問題 ................................................................................ 1 軟件估計(jì)常用方法 ....................................................................................... 3 風(fēng)險(xiǎn)管理過程框架 ....................................................................................... 5 常用的風(fēng)險(xiǎn)識別和風(fēng)險(xiǎn)評估方法 .................................................................. 7 本文的工作 ................................................................................................. 9 第二章 軟件項(xiàng)目的風(fēng)險(xiǎn)因子 ........................................................................................... 11 風(fēng)險(xiǎn)的定義 ................................................................................................ 11 風(fēng)險(xiǎn)的影響緯度 ......................................................................................... 11 風(fēng)險(xiǎn)的量化定義 .........................................................................................12 風(fēng)險(xiǎn)因子的定義 .........................................................................................14 軟件項(xiàng)目風(fēng)險(xiǎn)因子標(biāo)識方法 ........................................................................15 第三章 主要風(fēng)險(xiǎn)因子的潛在影響分析 .............................................................................17 實(shí)際軟件項(xiàng)目的風(fēng)險(xiǎn)因子標(biāo)識 ....................................................................17 主要風(fēng)險(xiǎn)因子原因結(jié)果圖 ...........................................................................19 風(fēng)險(xiǎn)因子影響調(diào)查 ......................................................................................25 風(fēng)險(xiǎn)因子影響圖曲線 ..................................................................................26 軟件主要風(fēng)險(xiǎn)因子對項(xiàng)目進(jìn)度的總體影響 ...................................................42 第四章 基于風(fēng)險(xiǎn)分析的軟件項(xiàng)目管理模擬模型 ...............................................................44 風(fēng)險(xiǎn)因子與不確定性 ..................................................................................44 軟件項(xiàng)目風(fēng)險(xiǎn)因子 ......................................................................................45 模擬模型 ....................................................................................................46 基于風(fēng)險(xiǎn)分析的軟件項(xiàng)目管理模擬模型介紹 ...............................................47 基于風(fēng)險(xiǎn)分析的軟件項(xiàng)目管理模擬模型的實(shí)現(xiàn) ............................................48 模擬模型使用案例 ......................................................................................52 模型驗(yàn)證 ....................................................................................................55 第五章 總結(jié)與展望 .........................................................................................................56 參考文獻(xiàn) .........................................................................................................................57 致謝 ................................................................................................................................59 南京大學(xué)碩士論文 1 第一章 緒論 本文研究的背景及問題 軟件已經(jīng)成為基于計(jì)算機(jī)的系統(tǒng)及產(chǎn)品成功的關(guān)鍵因素,其重要作用已經(jīng)得到了人們的普遍認(rèn)同。 本文實(shí)現(xiàn)了基于風(fēng)險(xiǎn)因子分 析的軟件項(xiàng)目管理模型,并對模型本身進(jìn)行了正確性驗(yàn)證,也在軟件項(xiàng)目組進(jìn)行了符合項(xiàng)目經(jīng)理需要的確認(rèn)。這需要綜合考慮軟件項(xiàng)目進(jìn)度計(jì)劃與軟件項(xiàng)目風(fēng)險(xiǎn)管理計(jì)劃,提供工具用以標(biāo)識、分析和管理軟件項(xiàng)目風(fēng)險(xiǎn),并在此基礎(chǔ)上獲得合理的軟件項(xiàng)目進(jìn)度計(jì)劃。能否在規(guī)定的時(shí)間內(nèi)交付軟件產(chǎn)品,與項(xiàng)目進(jìn)度計(jì)劃是否合理、項(xiàng)目風(fēng)險(xiǎn)管理活動(dòng)是否有效有很大的關(guān)系。同時(shí),利用模型中識別出的主要風(fēng)險(xiǎn)因子,標(biāo)識軟件項(xiàng)目風(fēng)險(xiǎn);綜合考慮風(fēng)險(xiǎn)因子的潛在影響和項(xiàng)目進(jìn)度的要求,制定出軟件項(xiàng)目風(fēng)險(xiǎn)管理計(jì)劃和合理的軟件項(xiàng)目進(jìn)度計(jì)劃。 南京大學(xué)碩士論文 3 ABSTRACT Many uncertainties are existed in software development process, and they give rise to risk of project success. Whether the project can deliver the product to the customer in time is much dependent on its estimated schedule plan and risk management plan. It is required to integrate software project schedule plan and software project risk management plan, and to offer tools for identifying, assessing, and managing the project risk, and to obtain a reasonable project schedule plan based on risk analysis. This paper has produced a software project management model based on risk factor analysis. Based on study of literatures and actual software projects developed in recent years of a famous munication pany, twenty risk factors that affect software project success are identified. The six main risk factors are selected and further quantitative analysis of their effects to project schedule is made. Monte Carlo method is used to simulate the total effects to project schedule, and the result is described as a risk graph. The project can identify project risk based on selected risk factors. By considering the potential effects of risk factors and the project schedule requirement, software risk plan and a reasonable software schedule plan can be made. A software project management model has developed in this paper. Model verificatio