本文關(guān)鍵詞： 灰色定權(quán)聚類方法 馬爾可夫預(yù)測模型 區(qū)間GERT網(wǎng)絡(luò) 白化權(quán)函數(shù) 概率矩陣　出處：《南京航空航天大學(xué)》2013年碩士論文　論文類型：學(xué)位論文

【摘要】：穩(wěn)定系統(tǒng)受沖擊作用的影響，其結(jié)構(gòu)和變動(dòng)方向?qū)?huì)發(fā)生變化。本文擬解決復(fù)雜大系統(tǒng)受外界沖擊作用，帶來的系統(tǒng)穩(wěn)定性變動(dòng)等變化，以系統(tǒng)的構(gòu)成元素為研究對象，通過將系統(tǒng)的各組成對象劃分不同的子系統(tǒng)，了解系統(tǒng)內(nèi)部構(gòu)造，針對不同子系統(tǒng)組成對象的特征分別構(gòu)建預(yù)測模型，預(yù)測系統(tǒng)的演化與發(fā)展方向。 本文共分為六章，通過層次劃分、模型構(gòu)建到結(jié)果預(yù)測，逐步深入了解系統(tǒng)結(jié)構(gòu)與發(fā)展趨勢。第二章構(gòu)建指標(biāo)體系將復(fù)雜的系統(tǒng)結(jié)構(gòu)條理化，完成對復(fù)雜系統(tǒng)的初步了解，實(shí)現(xiàn)將系統(tǒng)內(nèi)不同的指標(biāo)劃分為不同層次：區(qū)域性中心對象以及非區(qū)域性中心對象。第三、四章構(gòu)建灰色馬爾可夫預(yù)測模型，對區(qū)域性中心對象演化方向進(jìn)行預(yù)測。第五章構(gòu)建區(qū)間GERT網(wǎng)絡(luò)模型，對非區(qū)域性中心對象之間發(fā)生轉(zhuǎn)移的相關(guān)參數(shù)進(jìn)行預(yù)測。 本文共構(gòu)建三種模型：層次結(jié)構(gòu)劃分模型、馬爾可夫預(yù)測模型以及區(qū)間GERT網(wǎng)絡(luò)模型。在廣大學(xué)者研究的基礎(chǔ)上，，本文作者對上述三種模型的進(jìn)行局部改進(jìn)： （1）層次結(jié)構(gòu)劃分模型。對灰色白化權(quán)函數(shù)轉(zhuǎn)折點(diǎn)的選取方法進(jìn)行改進(jìn)，得到改進(jìn)轉(zhuǎn)折點(diǎn)的灰色定權(quán)聚類分析方法，篩選能夠體現(xiàn)系統(tǒng)對象的地位和功能指標(biāo)，建立指標(biāo)體系，構(gòu)建改進(jìn)轉(zhuǎn)折點(diǎn)的灰色定權(quán)聚類分析方法，劃分系統(tǒng)層次，驗(yàn)證改進(jìn)后的白化權(quán)函數(shù)性質(zhì)，設(shè)計(jì)模型的求解方法，實(shí)現(xiàn)對系統(tǒng)層次結(jié)構(gòu)的劃分。 （2）馬爾可夫預(yù)測模型。對傳統(tǒng)的馬爾可夫預(yù)測模型的概率轉(zhuǎn)移矩陣進(jìn)行改進(jìn)，將概率矩陣從實(shí)數(shù)矩陣擴(kuò)展到區(qū)間概率矩陣。通過對原始序列與預(yù)測序列的誤差值，確定不同狀態(tài)的區(qū)間范圍，將不同范圍內(nèi)的誤差值劃分為不同的狀態(tài)，得到指標(biāo)的區(qū)間概率矩陣，歸一化概率矩陣得到改進(jìn)后的灰色馬爾可夫模型。 （3）區(qū)間GERT網(wǎng)絡(luò)模型。傳統(tǒng)的GERT網(wǎng)絡(luò)的信息流參數(shù)均為實(shí)數(shù)，本文改進(jìn)GERT網(wǎng)絡(luò)信息流參數(shù)范圍，構(gòu)建網(wǎng)絡(luò)節(jié)點(diǎn)不同時(shí)間節(jié)點(diǎn)下的參數(shù)序列，考慮節(jié)點(diǎn)自身發(fā)生轉(zhuǎn)移以及節(jié)點(diǎn)間發(fā)生轉(zhuǎn)移的信息流變化情況，得到各參數(shù)信息流的區(qū)間序列，完成對區(qū)間GERT網(wǎng)絡(luò)模型構(gòu)建。 考慮滬寧高速鐵路線的開通對沿線城市經(jīng)濟(jì)發(fā)展的影響，通過第二章構(gòu)建的層次結(jié)構(gòu)模型劃分城市群的層次，將滬寧高速鐵路沿線城市群劃分為區(qū)域性中心城市和非區(qū)域性中心城市，第三、四章構(gòu)建的灰色馬爾可夫預(yù)測模型，預(yù)測了滬寧高速鐵路沿線區(qū)域性中心城市未來發(fā)展方向，第五章構(gòu)建的區(qū)間GERT網(wǎng)絡(luò)模型，預(yù)測了非區(qū)域性中心城市產(chǎn)業(yè)轉(zhuǎn)移的信息流參數(shù)。
[Abstract]:Under the influence of shock, the structure and direction of the stability system will change. This paper aims to solve the changes of the stability of the complex large scale system by the external shock, and take the component elements of the system as the research object. By dividing each component object of the system into different subsystems, the internal structure of the system is understood, and the prediction model is constructed according to the characteristics of the different subsystem components to predict the evolution and development direction of the system. This paper is divided into six chapters. By dividing the level, constructing the model to forecast the result, we can understand the system structure and the development trend step by step. In the second chapter, the index system will be constructed and the complex system structure will be organized, and the preliminary understanding of the complex system will be completed. The realization of the system will be divided into different indicators: regional central object and non-regional central object. In Chapter 5th, an interval GERT network model is constructed to predict the parameters related to the transition between non-regional central objects. In this paper, three kinds of models are constructed: hierarchical structure partition model, Markov prediction model and interval GERT network model. The method of selecting turning point of grey whitening weight function is improved, and the grey weight clustering analysis method of improved turning point is obtained, which can reflect the status and function index of system object, and establish the index system. The grey weight clustering analysis method of the improved turning point is constructed, the system hierarchy is divided, the properties of the improved whitening weight function are verified, the solution method of the design model is designed, and the division of the system hierarchy is realized. 2) Markov prediction model. The probability transfer matrix of the traditional Markov prediction model is improved, and the probability matrix is extended from the real number matrix to the interval probability matrix. The interval range of different states is determined and the error values in different ranges are divided into different states. The interval probability matrix of the index is obtained and the improved grey Markov model is obtained by normalizing the probability matrix. The information flow parameters of traditional GERT networks are all real numbers. This paper improves the range of information flow parameters of GERT networks and constructs the parameter sequences of network nodes under different time nodes. Considering the change of information flow between nodes and between nodes, the interval sequence of each parameter information flow is obtained, and the model of interval GERT network is constructed. Considering the impact of the opening of the Shanghai-Nanjing high-speed railway line on the economic development of the cities along the line, the hierarchical structure model constructed in the second chapter is used to divide the urban agglomeration. The urban agglomeration along the Shanghai-Nanjing high-speed railway is divided into regional central cities and non-regional central cities. In the third and fourth chapters, the grey Markov forecasting model is constructed to predict the future development direction of the regional central cities along the Shanghai-Nanjing high-speed railway. In chapter 5th, the interval GERT network model is constructed to predict the information flow parameters of industrial transfer in non-regional central cities.
【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2013
【分類號(hào)】：O211.62;F127

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 唐小我,王景,曹長修;一種新的模糊自適應(yīng)變權(quán)重組合預(yù)測算法[J];電子科技大學(xué)學(xué)報(bào);1997年03期

2 陳濤捷;馬爾可夫預(yù)測模型的SD解法[J];系統(tǒng)工程;1989年04期

3 鄭軼松;齊二石;鄭曉東;;基于圖示評(píng)審技術(shù)GERT的高科技產(chǎn)品開發(fā)研究[J];系統(tǒng)工程;2005年11期

4 阮愛清;劉思峰;;灰色GERT網(wǎng)絡(luò)及基于顧客需求的灰數(shù)估計(jì)精度[J];系統(tǒng)工程;2007年12期

5 劉名武;馬永開;;一類N-策略控制的Markov排隊(duì)制造系統(tǒng)質(zhì)量控制策略[J];系統(tǒng)工程;2009年12期

6 董博超;宋保維;梁慶衛(wèi);毛昭勇;;基于馬爾可夫模型的備件配置優(yōu)化方法[J];系統(tǒng)工程;2011年09期

7 劉芳;;基于灰色馬爾可夫模型的上海房地產(chǎn)業(yè)外商直接投資預(yù)測[J];經(jīng)濟(jì)視角(下);2008年07期

8 李紅梅;21世紀(jì)中國產(chǎn)業(yè)結(jié)構(gòu)調(diào)整的戰(zhàn)略選擇[J];首都師范大學(xué)學(xué)報(bào)(社會(huì)科學(xué)版);2000年06期

9 張浩;王福林;索瑞霞;申雷明;;基于馬爾可夫模型的我國東、中、西部的產(chǎn)業(yè)結(jié)構(gòu)預(yù)測[J];數(shù)學(xué)的實(shí)踐與認(rèn)識(shí);2010年14期

10 曹明霞;;南京市產(chǎn)業(yè)結(jié)構(gòu)調(diào)整的優(yōu)化模型與實(shí)證分析[J];統(tǒng)計(jì)與決策;2008年03期

本文編號(hào)：1533034