鱼群集群行为的建模与仿真.pdf

中文摘要群体智能是源于自然界群居生物系统中的一种涌现智能,是指许多简单个体通过相互合作产生复杂智能行为的特性。自然界中存在着大量群居生物系统,它们表现出高度的群体智能行为。群体智能强调个体的简单性,群体的涌现性以及自下而上的研究策略。生态系统中,动物个体的行为相对简单,但当它们集群后,却能表现出复杂的群体行为。目前对群体智能的研究仍处于初级阶段,研究和掌握群体智能系统的特性与规律,具有理论和应用两个方面的重要意义。本文选择了自然界的鱼群作为研究对象,并分析了鱼群的行为和特征,通过计算机仿真方法建立了一个简单的人工鱼系统,用图形化的方式展示了鱼群游动、捕食以及繁殖等行为,通过系统运行演示鱼群复杂行为产生的过程。本文首先建立了一种基于向量的数据模型,用来描述鱼群的行为规则。模型由环境、主体、行为规则组成。环境为海洋;主体包括两类具有捕食关系的鱼类主体、食物主体、鱼钩主体;行为规则包括游动规则、觅食规则、繁殖规则、死亡规则。基于以上规则,模型尽可能地反映鱼群的主要属性。其次,利用java技术设计了人工鱼系统,每个人工鱼遵守相同的行为规则交互和游动,最终涌现出复杂的群体行为。人工鱼能很好地模拟了真实鱼类的智能行为特征。最后,基于人工鱼系统,研究了鱼群复杂行为的产生及其原因。研究结果表明,群体智能的产生依赖于个体与环境以及个体之间的交互。关键词:群体智能;捕食;涌现;java;Agent The modeling and simulation oftheswarm behavior infishschool Graduate Name:Zhao puter Applied Technology) Directedby Zeng Jian—chao AB STRACT Swarm intelligence is a emerge intelligence stems from nature swarm intelligence is aquality thatmany simple individuals can plex intelligent behaviors through exist agreat many swarm ecosystems innature,performing highly swarm behavior. Swarm intelligence highlights the simplicity of individuals,emergent property ofthe colony and boaom-up research ecosystem, individual animals behave rather simply,while thecolony,once theyform, could show a plicated present,the research on swarm intelligence is still intheprimary stage,therefore,it isimportant to study and master theproperty and rules ofswarm intelligence system in terms oftheory andapplication. The dissertation chooses the natural fish astheobj ectiveofresearch,and analyzes thebehavior and characteristic offishschool,and aims toestablish asimple artificialfishsystem puter simulmion technology,and shows how fishcolony swim,prey on and breed by the graphical way. Through the run ofsystem,the process offishcolony’S plex behavior ,a vector-based datamodeliS constructed to describe the behavioral rules of model consias of and stands for ocean. The