本文選題：數(shù)控系統(tǒng)　切入點：軟件可靠性　出處：《合肥工業(yè)大學》2016年博士論文　論文類型：學位論文

【摘要】：數(shù)控機床是裝備制造業(yè)最重要的基礎機械,其整體性能在很大程度上是由數(shù)控系統(tǒng)的性能水平而決定的。近些年來,隨著我國機床產(chǎn)業(yè)的高速發(fā)展,國產(chǎn)數(shù)控系統(tǒng)的性能也得到了較大地提高,尤其是在功能和精度方面,已經(jīng)逐漸接近國外先進水平。但是,相關數(shù)據(jù)統(tǒng)計表明,我國大部分中高檔數(shù)控系統(tǒng)仍需依賴進口。究其原因,主要是國產(chǎn)數(shù)控系統(tǒng)在可靠性方面與國外相比仍存在較大差距。因此,提高國產(chǎn)數(shù)控系統(tǒng)的可靠性水平,已經(jīng)成為我國裝備制造業(yè)快速發(fā)展的關鍵所在。隨著硬件技術的不斷發(fā)展與完善,硬件平臺已經(jīng)非常成熟,其可靠性也相對較高,而軟件失效已經(jīng)成為導致整個系統(tǒng)發(fā)生故障的主要原因。由于加工精度和加工速度等技術指標的不斷提升,數(shù)控系統(tǒng)的軟件規(guī)模越來越大,軟件層次越發(fā)復雜,導致軟件發(fā)生失效的概率成倍增加。因此,研究數(shù)控系統(tǒng)軟件可靠性的相關技術和理論變得非常重要。本文結合國家科技重大專項“標準型數(shù)控系統(tǒng)的產(chǎn)業(yè)化及專用型齒輪機床數(shù)控系統(tǒng)的研究開發(fā)”(編號:2012ZX04001-021)和安徽省自主創(chuàng)新專項資金項目“高性能智能化數(shù)控系統(tǒng)開發(fā)”(編號:2013AKKG0394)的研究工作,針對數(shù)控系統(tǒng)軟件可靠性進行了全面而系統(tǒng)地研究,主要研究工作如下:(1)建立了數(shù)控系統(tǒng)軟件體系結構模型,主要分為功能層、程序層和組件層;建立了一個基于AHP和組件復雜度的數(shù)控系統(tǒng)軟件可靠性分配模型,以數(shù)控系統(tǒng)整體實用性最大為目標,以程序的最低可靠性和研發(fā)費用為約束條件,將可靠性指標分配給每個程序。并將程序的可靠性指標作為總指標,以組件的復雜度作為考量因素進行再分配,最終將可靠性指標分配給每個組件。實驗結果表明,該方法在數(shù)控系統(tǒng)設計前期,研發(fā)人員對程序中組件的調用情況掌握不準確的情況下,可以將數(shù)控系統(tǒng)軟件可靠性總指標合理地分配給每個組件,并為單個組件的研發(fā)提供了指導依據(jù)。(2)對GSK-HFUT型滾齒數(shù)控系統(tǒng)的軟件部分進行了全面的失效模式及影響分析(FMEA)。分析的目的是找出系統(tǒng)軟件中可靠性比較薄弱的環(huán)節(jié),從而采取一些改進措施以降低失效發(fā)生的概率。分別從系統(tǒng)級和詳細級兩個方面對該型滾齒數(shù)控系統(tǒng)進行了分析,分析結果對于提高系統(tǒng)的軟件可靠性或研發(fā)新系統(tǒng)具有一定的指導意義。(3)研究了數(shù)控系統(tǒng)軟件缺陷自定位方法,并建立了代碼缺陷懷疑度分類器模型。該方法可以幫助研發(fā)人員在軟件發(fā)生失效后迅速準確地對缺陷進行定位。在數(shù)控系統(tǒng)軟件體系結構的基礎上,增加了一個體系結構擴展層,用以跟蹤記錄軟件的執(zhí)行信息。通過所建立的數(shù)控系統(tǒng)軟件代碼缺陷懷疑度分類器,計算出失效路徑中各條有向邊的懷疑度,指導研發(fā)人員對懷疑度最高的有向邊進行代碼檢查。實驗結果表明該方法是可行的,可以有效提高故障排除效率,減少數(shù)控系統(tǒng)的平均修復時間,從而提高了數(shù)控系統(tǒng)的可用度。(4)建立了軟件可靠性模型評價體系和基數(shù)型多屬性決策模型。通過6個評價指標,分別從擬合能力和預測能力兩方面對模型進行全面地定量地評價。結合多目標決策方法對軟件可靠性模型進行綜合評價與選擇,并討論了決策模型的3種求解方法。實驗結果表明,利用基數(shù)型多屬性決策模型來選擇最優(yōu)軟件可靠性模型是可行的。(5)建立了數(shù)控系統(tǒng)軟件可靠性評估與預測模型。通過加權組合的建模方式來提高模型預測的準確性。提出了動態(tài)權值調整算法,并建立了動態(tài)加權組合模型。對比實驗表明,該模型比現(xiàn)有的模型具有更低的擬合誤差。在GSK-HFUT型滾齒數(shù)控系統(tǒng)上的成功應用,證明了此建模方法是有效的、可行的和普遍適用的,具有非常高的實用價值,其預測結果對于管理層作出正確決策具有非常重要的意義。
[Abstract]:NC machine tool is the basis of mechanical equipment manufacturing industry is the most important, its overall performance is determined by the performance level of CNC system to a great extent. In recent years, with the rapid development of China's machine tool industry, the performance of domestic CNC system have been greatly improved, especially in the aspects of function and accuracy, has been gradually close to the advanced level. However, the relevant statistical data show that, in the most high-end CNC system in China is still dependent on imports. The reason is mainly domestic CNC system in terms of reliability compared with foreign countries there is still a large gap. Therefore, improving the reliability level of domestic CNC system, has become the key to the rapid development of equipment manufacturing industry of our country. With the development of hardware technology, the hardware platform has been very mature, its reliability is relatively high, and has become the cause of software failure The main reason of failure of the system. The technical indexes of machining precision and machining speed increase, the CNC system software is more and more large scale software more complex, causes the software failure probability is exponentially increasing. Therefore, the related technology and theory research of CNC system software reliability becomes very important. Based on research and development industrialization and special type gear CNC system standard CNC system of the major projects of national science and technology "(No. 2012ZX04001-021) and the independent innovation in Anhui Province special fund project" high performance development of intelligent CNC system "(serial number: 2013AKKG0394) of the research work, the NC system software reliability research comprehensively and systematically are the main research work is as follows: (1) the establishment of the CNC system software architecture model, mainly consists of function layer, application layer and component layer; The establishment of a AHP and component based on complex CNC system software reliability allocation model, with the goal of maximum overall utility of CNC system, with the lowest cost of research and development program reliability as the constraint condition, the reliability index is assigned to each program. And the program reliability indexes as the total index, the complexity of the components as consideration for redistribution, finally the reliability index assigned to each component. The experimental results show that the method in the numerical control system design, R & D personnel master calls to the program components in the case of inaccurate reliability of CNC system software can be reasonably allocated to the total index of each component, and provides guidance on the basis of for the development of individual components. (2) the part of the software of GSK-HFUT hobbing CNC system of failure mode and effect analysis (FMEA) comprehensive analysis of the objective. Is to find out the weak links on the reliability of the software of the system, in order to take some measures to reduce the probability of failure. From two aspects of system level and detail level of the hobbing CNC system is analyzed, the analysis result has certain guiding significance for improving the reliability of software system or the development of new systems (3.) studied self localization method of CNC system software defects, and set up the code defect doubt classifier model. This method can help the developers in the software after the failure occurred quickly and accurately the defect location. Based on the numerical control system software architecture, adds an architecture extension layer, in order to track record the software execution information. Through the numerical control system software code defects the doubt classifier, calculate the failure path to the edge of the doubt, guide the development of people To doubt the highest degree of a code checks to the edge. The experimental results show that this method is feasible and can effectively improve troubleshooting efficiency, reduce the average repair time of NC system, so as to improve the availability of the CNC system. (4) to establish the model of software reliability evaluation system and multi attribute decision making model base through the 6 evaluation indexes, respectively from the fitting ability and predictive ability of the two aspects of the model to conduct a comprehensive quantitative evaluation. Combining the multi-objective decision method is used to evaluate the software reliability model selection, and discusses 3 kinds of methods for solving decision-making model. Experimental results show that the base of multi attribute decision making model optimal selection of software reliability model is feasible. (5) the establishment of the evaluation and prediction model of software reliability of CNC system. Through modeling the weighted combination to improve the model prediction accuracy . the dynamic weight adjustment algorithm, and establishes a dynamic weighted combination model. Experiments show that the fitting error of the model is lower than the existing model. The successful application of CNC gear rolling system in GSK-HFUT, proved that this modeling method is effective, feasible and applicable, with very high the practical value of the prediction results for management decision making has very important significance.

【學位授予單位】：合肥工業(yè)大學
【學位級別】：博士
【學位授予年份】：2016
【分類號】：TG659;TP311.5

【相似文獻】

相關期刊論文 前10條

1 黃金慶，，劉明烈;高性能數(shù)控系統(tǒng)軟件劃分與調度[J];制造技術與機床;1995年11期

2 范菲,陳幼平,胡文彬,周祖德;數(shù)控系統(tǒng)軟件芯片庫集成開發(fā)環(huán)境的研究[J];華中理工大學學報;1998年02期

3 王軼辰;數(shù)控系統(tǒng)軟件的測試與故障診斷[J];設備管理與維修;2005年02期

4 喬巍巍;賈亞洲;王桂萍;張學文;盛菲;;數(shù)控系統(tǒng)軟件故障可靠性分析[J];制造技術與機床;2006年10期

5 黃興;唐小琦;周向東;;總線型數(shù)控系統(tǒng)軟件平臺的研究與實現(xiàn)[J];組合機床與自動化加工技術;2010年12期

6 謝明紅;開放式模塊化通用數(shù)控系統(tǒng)軟件設計[J];華僑大學學報(自然科學版);1998年02期

7 左靜,魏仁選,呂新平,陳幼平,周祖德;數(shù)控系統(tǒng)軟件芯片的研制和開發(fā)[J];中國機械工程;1999年04期

8 胡建中,周云飛,李左章;數(shù)控系統(tǒng)軟件任務劃分與調度[J];機械與電子;1999年02期

9 李芬;常勇;王威;;基于嵌入式linux的數(shù)控系統(tǒng)軟件設計[J];制造業(yè)自動化;2012年09期

10 蔣知峰;湯季安;;打造自主品牌的高端數(shù)控系統(tǒng)軟件平臺[J];上海信息化;2012年07期

相關會議論文 前1條

1 齊智平;;具有開放式體系結構的數(shù)控系統(tǒng)軟件平臺的研究與實現(xiàn)[A];西部大開發(fā) 科教先行與可持續(xù)發(fā)展——中國科協(xié)2000年學術年會文集[C];2000年

相關博士學位論文 前3條

1 王力超;數(shù)控系統(tǒng)軟件可靠性及評估方法研究與應用[D];合肥工業(yè)大學;2016年

2 谷巖;數(shù)控系統(tǒng)軟件可靠性設計與故障分析技術[D];吉林大學;2014年

3 黃海鵬;多軸聯(lián)動電火花加工數(shù)控系統(tǒng)軟件構建及其應用研究[D];哈爾濱工業(yè)大學;2010年

相關碩士學位論文 前10條

1 盛菲;數(shù)控系統(tǒng)軟件可靠性研究及應用[D];吉林大學;2004年

2 林桂清;基于文本指令的嵌入式數(shù)控系統(tǒng)軟件開發(fā)[D];中國地質大學;2009年

3 田軍鋒;基于Windows CE數(shù)控系統(tǒng)軟件的設計與實現(xiàn)[D];中國科學院研究生院（沈陽計算技術研究所）;2009年

4 譚菲菲;基于uTenux的數(shù)控系統(tǒng)軟件的研究與實現(xiàn)[D];大連理工大學;2013年

5 凌澤斌;慢走絲線切割數(shù)控系統(tǒng)軟件的研究[D];南昌大學;2008年

6 荀靖玢;基于嵌入式L的數(shù)控系統(tǒng)軟件的研究與開發(fā)[D];天津大學;2012年

7 戴乾坤;嵌入式數(shù)控系統(tǒng)軟件設計及其關鍵算法研究[D];中南大學;2008年

8 任參;數(shù)控系統(tǒng)軟件安全與插補算法研究[D];華南理工大學;2010年

9 丁仕燕;基于PC的數(shù)控系統(tǒng)軟件設計研究[D];合肥工業(yè)大學;2002年

10 李丁;基于ServoWorks的雕銑機數(shù)控系統(tǒng)軟件研究與開發(fā)[D];中國科學院研究生院（沈陽計算技術研究所）;2008年

本文編號：1593138