本文選題：人類動力學(xué) + 車載自組網(wǎng)��； 參考：《吉林大學(xué)》2016年博士論文

【摘要】：人類動力學(xué)在互聯(lián)網(wǎng)中的廣泛存在現(xiàn)象已經(jīng)被越來越多的學(xué)者研究和證實,各種網(wǎng)絡(luò)系統(tǒng)中的非泊松現(xiàn)象被挖掘分析,基于動力學(xué)的網(wǎng)絡(luò)模型也逐漸被建立。隨著移動網(wǎng)絡(luò)尤其是繼智能手機之后的車載移動通訊的出現(xiàn)和快速發(fā)展,使得智能車聯(lián)網(wǎng)絡(luò)成為了新的網(wǎng)絡(luò)研究熱點,并且在其他網(wǎng)絡(luò)系統(tǒng)中所發(fā)現(xiàn)的動力學(xué)因素是否同樣存在并且作用于車聯(lián)網(wǎng)中,以及以一種什么樣的形式存在且影響著未來道路交通的狀況值得關(guān)注。本文致力于挖掘車聯(lián)網(wǎng)中的動力學(xué)因素,利用人類動力學(xué)知識和方法優(yōu)化和建立智能車輛網(wǎng)絡(luò)中的相關(guān)問題模型。重點分析了車輛可通訊的未來道路與傳統(tǒng)道路的特性和不同,利用動力學(xué)方法對通訊方式進行了優(yōu)化,并且根據(jù)人類動力學(xué)知識建立了新的無人駕駛車輛駕駛員的模型。主要貢獻包括:1.分析了車輛網(wǎng)絡(luò)中社會屬性因素。通過對車輛網(wǎng)絡(luò)中的相關(guān)概念進行定義,對三組公開車輛行駛信息數(shù)據(jù)分別進行建模、比較,通過社會網(wǎng)絡(luò)的方法,分析車聯(lián)網(wǎng)絡(luò)中的社會行為屬性。2.建立了智能駕駛和傳統(tǒng)駕駛的車輛移動模型,比較了二者產(chǎn)生的道路拓?fù)涞木W(wǎng)絡(luò)屬性。根據(jù)智能駕駛的廣視野性、預(yù)先判斷和提前規(guī)劃的特性,建立了基于智能駕駛和傳統(tǒng)駕駛的兩種車輛移動模型;進一步對兩種移動模型進行同樣道路的仿真實驗、對所產(chǎn)生的道路拓?fù)溥M行了對比和分析,試圖發(fā)現(xiàn)兩種道路拓?fù)涞牟煌?并且通過移動模型的建立找出道路拓?fù)渚W(wǎng)絡(luò)屬性不同的原因。3.提出了兩種基于動力學(xué)的車聯(lián)網(wǎng)通訊優(yōu)化方法。(1)基于動力學(xué)的分簇方法:根據(jù)駕駛?cè)说纳鐣P(guān)系預(yù)先產(chǎn)生基于社會關(guān)系的駕駛員分組,進一步在行駛過程中,結(jié)合實時的位置和速度關(guān)系,產(chǎn)生車輛通訊的新的分簇方法;(2)基于動力學(xué)的跨層聯(lián)合優(yōu)化算法:在車輛通訊過程中,根據(jù)不同的利益要求,在引入容錯機制后,根據(jù)觀測動態(tài)地調(diào)節(jié)不同網(wǎng)絡(luò)層的參數(shù)設(shè)置,并且通過群體博弈方法使得道路整體利益得到最大化,而不是使得車輛節(jié)點個體利益得到最大化。4.建立了基于人類動力學(xué)的駕駛員模型。通過分析影響駕駛員駕駛行為的動力因素,對該過程中的人類行為因素進行分類提取,計算出駕駛員動力學(xué)因子,進一步使該行為因子作用于駕駛過程中的四個子過程(這里,我們首次把駕駛過程分解為四個獨立的動作,駕駛過程被認(rèn)為是這四個動作的聯(lián)合發(fā)生或者單獨執(zhí)行的結(jié)果),從而建立基于人類動力學(xué)的駕駛員模型。
[Abstract]:The widespread phenomenon of human dynamics in the Internet has been studied and verified by more and more scholars. The non-Poisson phenomenon in various network systems has been excavated and analyzed, and the dynamics based network model has been gradually established.With the emergence and rapid development of mobile network, especially after the smart phone, the smart car network has become a new research hotspot.And whether the dynamic factors found in other network systems also exist and act on the vehicle network, and what kind of form exist and affect the future road traffic situation is worthy of attention.This paper is devoted to excavating the dynamic factors in vehicle networking, optimizing and establishing the relevant problem models in intelligent vehicle network by using human dynamics knowledge and methods.In this paper, the characteristics and differences between the future road and the traditional road are analyzed, and the communication mode is optimized by means of dynamic method, and a new model of driverless vehicle driver is established according to the knowledge of human dynamics.The main contributions include: 1.The social attribute factors in vehicle network are analyzed.Through the definition of the related concepts in the vehicle network, three groups of open vehicle driving information data are modeled and compared, and the social behavior attributes of the vehicular network are analyzed by the method of social network.The vehicle moving models of intelligent driving and traditional driving are established, and the network properties of the road topology generated by the two models are compared.According to the characteristics of wide vision, prejudgment and advance planning of intelligent driving, two kinds of vehicle moving models based on intelligent driving and traditional driving are established, and the simulation experiments of the two kinds of moving models are carried out on the same road.By comparing and analyzing the road topology, we try to find out the difference between the two kinds of road topology, and find out the reason why the road topology network attribute is different by establishing the moving model.This paper proposes two dynamic optimization methods for vehicle networking communication. (1) A dynamic clustering method is proposed. According to the driver's social relations, the driver groups based on social relations are generated in advance, and further in the driving process,Combined with the real-time position and speed relationship, a new clustering method for vehicle communication is proposed. A dynamic cross-layer joint optimization algorithm is proposed: in the process of vehicle communication, according to different interests, after introducing fault-tolerant mechanism,According to the observation, the parameters of different network layers are dynamically adjusted, and the overall interests of the road are maximized by the group game method, rather than the individual interests of the vehicle nodes maximized. 4.The driver model based on human dynamics is established.By analyzing the dynamic factors that affect driver's driving behavior, the human behavior factors in the process are classified and extracted, and the driver's dynamic factors are calculated, which further make the behavioral factors act on the four sub-processes in the driving process.
【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2016
【分類號】：U495;TN929.5

【參考文獻】

相關(guān)期刊論文 前7條

1 郭進利;;博客評論的人類行為動力學(xué)實證研究和建模[J];計算機應(yīng)用研究;2011年04期

2 樊超;郭進利;紀(jì)雅莉;高蕾;張偉;;基于圖書借閱的人類行為標(biāo)度律分析[J];圖書情報工作;2010年15期

3 郭進利;;一個人類行為動力學(xué)模型及其精確解[J];物理學(xué)報;2010年06期

4 郭進利;;關(guān)于記憶對人類動力學(xué)影響的討論[J];上海理工大學(xué)學(xué)報;2010年03期

5 趙庚升;張寧;周濤;;網(wǎng)頁瀏覽中的標(biāo)度行為研究[J];統(tǒng)計與決策;2009年01期

6 李楠楠;張寧;周濤;;人類通信模式中基于時間統(tǒng)計的實證研究[J];復(fù)雜系統(tǒng)與復(fù)雜性科學(xué);2008年03期

7 張寧;;群體興趣網(wǎng)的統(tǒng)計特性研究[J];上海理工大學(xué)學(xué)報;2008年03期

，

本文編號：1744130