本文選題：信息物理融合系統(tǒng) + 信任網(wǎng)絡(luò)�。� 參考：《昆明理工大學(xué)》2016年碩士論文

【摘要】：隨著信息技術(shù)的研究和互聯(lián)網(wǎng)產(chǎn)業(yè)的快速發(fā)展,直接推動(dòng)了信息物理融合系統(tǒng)(Cyber-physical systems, CPS)的開展與應(yīng)用。CPS依托云計(jì)算、大數(shù)據(jù)、通信網(wǎng)絡(luò)、嵌入式等多領(lǐng)域技術(shù),實(shí)現(xiàn)了物與物之間的高度融合與集成,其中信任關(guān)系作為網(wǎng)絡(luò)環(huán)境中非常重要的影響因素是值得關(guān)注和研究的。CPS作為互聯(lián)網(wǎng)的一種延伸,網(wǎng)絡(luò)環(huán)境既有相同也有不同之處。目前大部分信任模型主要研究范圍是在互聯(lián)網(wǎng)環(huán)境下,基于CPS應(yīng)用環(huán)境下的信任模型很少涉及,本文將基于CPS應(yīng)用環(huán)境下對(duì)信任網(wǎng)絡(luò)進(jìn)行研究。本文首先研究了互聯(lián)網(wǎng)下集中式和分布式兩大現(xiàn)有的主要的信任網(wǎng)絡(luò)模型,分析了兩種信任模型的特點(diǎn)和應(yīng)用場(chǎng)景,結(jié)合CPS的具體環(huán)境,構(gòu)建了CPS下中心控制型,鏈路相關(guān)型和中心相關(guān)型3種信任網(wǎng)絡(luò)模型。中心控制型信任網(wǎng)絡(luò)模型的應(yīng)用場(chǎng)景主要是在信任機(jī)構(gòu)進(jìn)行主導(dǎo)評(píng)價(jià)的環(huán)境中。在信任評(píng)價(jià)過程中結(jié)合CPS環(huán)境提出了基于SERVQUAL模型的信任評(píng)估考量因素,考慮了評(píng)價(jià)數(shù)據(jù)的客觀性和主觀性的差異,引入了D-S融合理論消弱主觀因素對(duì)信任評(píng)價(jià)結(jié)果的影響,最后對(duì)主觀評(píng)價(jià)結(jié)果和客觀評(píng)價(jià)結(jié)果進(jìn)行加權(quán)處理給出最后的信任評(píng)價(jià)值,并通過實(shí)例驗(yàn)證了方法的可行性。鏈路相關(guān)型信任網(wǎng)絡(luò)模型的應(yīng)用場(chǎng)景主要是在網(wǎng)絡(luò)中節(jié)點(diǎn)與節(jié)點(diǎn)之間已存在相當(dāng)數(shù)量的信任評(píng)價(jià)關(guān)系的環(huán)境中,需求節(jié)點(diǎn)和目標(biāo)節(jié)點(diǎn)可以通過其它節(jié)點(diǎn)來(lái)形成兩個(gè)節(jié)點(diǎn)之間的信任價(jià)值鏈。本文在對(duì)信任關(guān)系的研究基礎(chǔ)上給出了信任價(jià)值鏈的優(yōu)化規(guī)則,通過優(yōu)化計(jì)算最終得到需求節(jié)點(diǎn)和目標(biāo)節(jié)點(diǎn)信任評(píng)估值,并通過實(shí)例驗(yàn)證了方法的可行性。中心相關(guān)型信任網(wǎng)絡(luò)模型的應(yīng)用場(chǎng)景主要是已有不同節(jié)點(diǎn)對(duì)中心節(jié)點(diǎn)有過評(píng)價(jià)歷史數(shù)據(jù),新的引入節(jié)點(diǎn)通過評(píng)價(jià)歷史數(shù)據(jù)確定對(duì)中心節(jié)點(diǎn)的信任評(píng)價(jià)值。本文以具體應(yīng)用環(huán)境為例,對(duì)信任過程進(jìn)行優(yōu)化計(jì)算,并通過實(shí)例驗(yàn)證了方法的優(yōu)越性。
[Abstract]:With the research of information technology and the rapid development of Internet industry, the development and application of Cyber-physical systems (CPS), a physical information fusion system, rely on cloud computing, big data, communication network, embedded and other multi-domain technologies. The integration of objects and objects is realized, and the trust relationship is an extension of the Internet, which is a very important influence factor in the network environment. The network environment has some similarities and differences. At present, most of the trust models are mainly studied in the Internet environment, but the trust model based on CPS application environment is rarely involved. This paper will study the trust network based on CPS application environment. In this paper, the two main trust network models under the Internet, centralized and distributed, are studied, and the characteristics and application scenarios of the two trust models are analyzed. Combined with the specific environment of CPS, the central control model under CPS is constructed. There are three kinds of trust network models: link correlation and central correlation. The application scenario of the central control trust network model is mainly in the environment of trust organization leading evaluation. In the process of trust evaluation, considering the difference of objectivity and subjectivity of evaluation data, considering the factors of trust evaluation based on SERVQUAL model, combining with CPS environment, the D-S fusion theory is introduced to weaken the influence of subjective factors on the results of trust evaluation. Finally, the final trust evaluation value is given by weighting the subjective evaluation results and the objective evaluation results, and the feasibility of the method is verified by an example. The application scenario of the link-dependent trust network model is mainly in the environment where there are a number of trust evaluation relationships between nodes in the network. The requirement node and the target node can form the trust value chain between the two nodes through other nodes. Based on the research of trust relationship, this paper gives the optimization rules of trust value chain, and finally obtains the trust evaluation value of requirement node and target node by optimizing calculation. Finally, the feasibility of the method is verified by an example. The application scenario of the center-dependent trust network model is that different nodes have evaluated the historical data of the central node, and the new introduction node determines the trust evaluation value of the central node by evaluating the historical data. This paper takes the concrete application environment as an example, carries on the optimization computation to the trust process, and verifies the superiority of the method through the example.
【學(xué)位授予單位】：昆明理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP393.4;TP202
，

本文編號(hào)：1830213