摘要
一直以来，三维视景仿真基本上是在单机或者小范围局域网环境下进行的。随着计算机网络通讯技术和图形图像技术的发展，使得在分布式环境下进行三维视景仿真成为可能。
分布式交互三维视景仿真平台就是利用现有的网络条件，为大量的计算机共同完成三维视景仿真任务提供一个较为通用的平台。它为在分布式三维视景仿真环境下遇到的一些具有共性的问题提供一套解决方案。平台能够降低仿真开发的难度，使仿真开发人员能够将精力主要集中在解决具体仿真领域的算法上，这样缩短仿真开发时间，提高了生产效率。因此，2000年国家自然科学基金项目“分布交互三维视景行为—特征建模方法研究”展开了对3D建模技术、3D碰撞检测、人体运动仿真、分布式网络通讯模型、3D寻径算法和数据库引擎技术等研究开发，力图开发出一个应用面较广的仿真平台。
本文主要结合该课题中有关分布式网络通讯模型、3D寻径算法和数据库引擎技术进行了探讨；以A~*、B+Tree、TCP／IP和LALR(1)等理论技术为基础，对以下几个问题进行了研究并提供了相应解决方案：如何有效利用网络带宽提高数据交换的质量，降低网络延时对系统仿真的影响；在分布式环境下虚拟物体如何实现快速地寻径；仿真数据如何被有效的存储和读取。本文对分布式系统的开发有着一定的理论和实用价值。
第1章 绪论，介绍了分布式交互三维视景仿真平台的应用背景，国内外相关的研究动向以及系统的框架结构。
第2章 分布式环境下的网络通讯技术，主要讨论了数据交换模型、通讯协议的选择以及分布式系统的同步等几个方面的问题。
第3章 虚拟环境下的快速3D寻径，主要讨论了人工智能中的A~*算法以及如何通过该算法实现三维环境下的快速寻径。
第4章 微型数据库引擎，在详细讨论了B+Tree、LALR(1)算法的基础之上，列出了数据存储的模型和数据库信息查询的机制。
第5章 通讯模块、快速寻径和微型数据库引擎的实现，介绍了这几个模块的框架结构以及完成的状况。
武汉理工大学硕士学位论文
第6章研究工作总结，主要对所开发的系统作出总结，指出了不足之
处，并展望了未来各项技术的发展。
Up to now, three-dimension visual simulation is basically implemented on a single computer or within a LAN that covers small area. The development of the computer network communication technology and graphic & image technology makes the implementing three-dimension scenery simulation under distributed environment feasible.
Three-dimension scenery simulation platform, utilizing the existing network, provides a comparative general platform for a large number of computers to cooperate to fulfill the three-dimension scenery emulation work. It affords a set of resolution schema for problems having some in common that will be confronted under the distributed three-dimension scenery emulation environment. The platform can release developers from some problems during the course of the developing and allows the developers to concentrate on the algorithms in the domain of concrete simulation, which leads to shorter developing time and more efficiency. Thus, the Nation Science Foundation Project- "Research of the modeling method for distributed interactive three-dimension visual behavior-characteristic", in 2000-carries out researches in3D modeling technology, 3D collision testing, human motion simulation, distributed network communication model, 3D routing algorithm and database engine technology, striving for developing a simulation pl
atform for general applications.
The article references the distributed network communication model, 3D routing algorithm and database engine technology which are parts of the project, based on A*, B+Tree, TCP/IP and LALR(l) , discusses the following problems and proposes resolutions: how to make efficient use of network bandwidth to promote the quality of data exchange and impair the impact of the network delay to the simulation system; how the virtual object can route rapidly under the distributed environment; how the data of simulation can be
.
stored and be read validly. The article is of great value in the theory and practice of the distributed system development.
Chapter 1: Preface. This chapter introduces the implementation background of distributed interactive three-dimension scenery simulation platform, domestic and foreign research trend in the related fields as well as architecture of the system.
Chapter 2: The network communication technology under the distributed environment. Mainly discusses the data exchange model, selection of communication protocol as well as the synchronization in distributed system.
Chapter 3: Rapid 3D routing in the virtual environment. Discusses the A* algorithm in the artificial intelligence field and how to use the algorithm to carry out rapid routing in 3D environment.
Chapter 4: Micro database engine. Proposes data store model and database information querying mechanism after detailed analysis of B+Tree and LALR(1) algorithm.
Chapter 5: The realization of communication, rapid 3D routing and micro database engine. Introduce the frame work and the progress of these components.
Chapter 6: Conclusion of the research. Comes to the conclusion of the developed system, points out insufficiencies and predicts the prospect of various technologies in the future.
引文
[1]．尹勇，金一丞，李志华．分布式航海仿真系统中的网络通讯．系统仿真学报．2000．12(6)：621～624
[2]．张新艳，张煜，王少梅等．基于虚拟现实的港口集装箱码头装卸系统仿真建模技术．武汉理工大学学报(交通科学与工程版)．2001．25(4)470～473
[3]．鲍广宇，刘晓明．通用分布式环境研究．解放军理工大学学报(自然科学版)．2003．4(1)：21～25
[4]．徐佑军，吴元保．基于WinSock的网络实时通信程序的设计．计算机工程．1999．25(1])：98～100
[5]．吴玲达，张茂军．分布式虚拟环境．计算机工程与应用．2002．7(8)：119～122
[6]．何进，谢松巍．基于Socket的TCP／IP网络通讯模式研究．计算机应用研究．2001．8(1)：134～135
[7]．谢安明，邵美珍．使用Socket设计网络通讯程序．信息工程大学学报．2001．2(1)：68～70
[8]. K.Sundaraj,C.Laugier, I.F.Costa. An approach to LEM modeling: Construction,Collision Detection and Dynamic Simulation. IEEE/RSJ Int. Conf. on Intelligent Robots and Systems. 2002.3(13): 1～6
[9]．杨克俭，刘舒燕，陈定方．分布交互三维视景行为．特征建模方法研究．计算机辅助设计与图形学学报．2000．12(11)：846～850
[10]．清宏计算机工作室．Visual C++编程技巧(网络与数据库篇)．北京：机械工业出版社．2001．134～267
[11]．李冠一，郑辉，韩维桓．基于WinSock实现Internet协议多线程连接的关键技术．计算机应用．2001．21(12)：50～52
[12]. Stankovic N,Zhang K. A Distributed Parallel Programming Framework[J]. IEEE transactions on software engineering .2002,28 (5):478～493.
[13]．刘培志，周卫华，樊利霞，吴江．分布式仿真中实时网络通讯技术研究．计算机仿真．2000．17(5)：48～52
[14]．杨克俭．基于分布式虚拟现实技术的船舶航行协同仿真系统研究．交通与计算机．2000．5(18)：29～32
[15]. Alice E. Fischer. Finding the Mid-Path. Wuhan University Journal of Natural Sciences. 2001.6(1):83～99
[16]. Liu Guizhen,Zhan Lanju. FRACTIONAL (g,f)-FACTORS OF GRAPHS. ACTA MATHEMATICA SCIENTIA. 2001.21(4):541～545
[17]. Liu Chunlin, He Jianmin, Shi Jianjun. New Methods to Solve Fuzzy Shortest Path Problems. Journal of Southeast University. 2001. 17(1):18～21
[18]．胡玉兰，朱立忠．基于遗传算法的机器人运动路径规划的应用研究．机械设计与制造．2002．5(2)：109～112
[19]．景兴建，王越超．一种基于理性遗传算法(RGA)的协调运动行为合成算法．机器人．2002．24(1)：49～55
[20]．孟庆浩，彭商贤，刘大维．基于Q-M图启发式搜索的移动机器人全局路径规划．计算机应用研究．1998．20(4)：273～279
[21]．王孙安，庄健．移动机器人路径发现与规划的免疫算法．系统仿真学报．2002．14(8)：995～997
[22]. Hashem M M A, Keigo Watanabe, Kiyo taka Izumi. An Evolutionary Optimal Obstacle Avoidance Method for Mobile Robots. Proc of The Fourth Int Symp on Artificial Life and Robotics(AROB 4th'99) ,O ita Japan, 1999, 618～621
[23]. LIU Yanfei,Qiu Yuhuang. Robot Path Planning Based on Genetic Algorithms with Two-Layer Encoding. CONTROL THEORY AND APPLICATIONS. 2000.17(3):429～432
[24]. Rghu Ramakrishnan,Johannes Gehrke. Database Manamengent Systems[M].北京：清华出版社．2002．1～422
[25]．陈仲民，王雪．面向对象数据库的索引技术．计算机工程与科学．2000．22(6)：100～104
[26]．董佑平，彭渤．SQL Server索引技术．山东交通学院学报．2004．12(1)：68～71
[27]．Hector Garcia-Molina,Jeffrey D．Ullman,Jennifer Widom．Database System Implentation．北京：机械工业出版社．2002．1～122
[28]．左为平．一类特殊的平衡二叉树．天水师范学院学报．2003．23(5)：16～17
[29]．赵连伟，罗四维等．二叉树动态负载平衡方法．计算机应用．2003．23(7)：52～53
[30]. Robinson J.T.The K-D-B-Tree: A Search structure for Large Multidimensional Dynamic Indexes [C]. In:Proc.ACM-SIGMOD.International conference on management of Data ,Ann Arbor ,April/may 1981. 10～18.
[31]．丁宇新，程虎．Java虚拟机中无用单元的精确回收．计算机学报．1999．22(11)：1228～1232
[32]．孙希，俞建新．Java虚拟机新技术探讨．计算机应用．2001．21(8)：74～76
[33]. Ole Agesen. Garbage collection and local variable type-precisionand liveness in Java virtual machine. In: Proc of SIGPLAN '98 Conference on Programming Languages Design and Implementation, Atlanta, 1998. 269～279
[34]．罗永龙，黄刘生．散列表中双重HASH函数的设计与分析．计算机工程与应用．2002．12(9)：59～60
[35]．任世军，杨姗姗．路径规划中两个多面体的快速碰撞检测算法．哈尔滨工业大学学报．2001．33(2)：252～254
[36]．陈晓波，熊光楞，柴旭东．基于分布式仿真模型对象的分布式协同建模技术研究．系统仿真学报．2002．14(5)576～580
[37]．孟庆浩，张明路，刘大维，彭商贤．基于双向A3算法的自主车全局路径规划．天津大学学报．1998．31(6)：747～751
[38]. Keymeulen D,Decuyper J. The Fluid Dynamics applied to Mobile Robot Motion:the Stream Field Method [A]. 1994 IEEE International Conference on Robotics and Automation. San Diego California: sponsored by IEEE Robotics and Automation Society, 1994. 378～385.
[39]. WAN G Guangrong,GU N aijie. An Efficient Parallel Minimum Spanning Tree Algorithmon Message Passing Parallel Machine. Journal of Software. 2000. 11(7):889～898
[40]. Boehm. Garbage collection in an uncooperative environment. Software Practice and Exp. 1988.18(9): 807～820
[41]．吴丽娟 徐心和．基于遗传算法的足球机器人比赛中障碍回避策略的设计．机器人．2001．23(2)：142～145
[42]. Fujisawa K,Hayakawa S, Aok i T. Real-Time Search for Autonomous Mobile Robot Using the Frameword of Anytime Algorithm. Proc of The Fourth Int Symp on Artificial Life and Robotics (AROB 4th'99) ,O ita Japan, 1999, 291～296
[43]．胡九龙，赵捧未．数据检索中索引技术研究．科技情报开发与经济．2004．14(1)：210～211
[44]．杨增芳．空间数据索引结构及聚类方法．玉溪师范学院学报．2002．18(5)：106～108
[45]. Elisa Bertino, Beng Chinooi. The Indispensability of Dispensable Index [J].IEEE Transaction on knowledge and DataEngineering. 1999.11(1):17～26.
[46]. T.Suganuma, et al. Overview of the IBM Java Just-in-TimeCompiler[J]. IBM Systems Journal, 39(1) .