Quantitative Simulation And Analysis Of Complex Adaptive System

There is great need in research on the problems of complex system and complexity along with entering into the 21st century. These problems are concerned with the human survival and the sustainable development of the society, including the problems of environment, resources, economy, population, health, disaster, peace and security, etc. The interdisciplinary collaboration and integration in a wholeness perspective must be adopted to solve these problems.Based on the complex adaptive system theory, multi-agent modeling and simulation provides a new approach to study complex adaptive systems and becomes one of the most popular approaches in this domain. According to the requirements of large scale complex adaptive system simulation, this thesis studied the key technology of simulation and analysis of complex adaptive system quantitatively based on multi-agent approach.Firstly, a overview of some basic concepts and their quantitative research status is introduced and discussed. Then, several concept pairs which easy to be confused are identificated respectively, providing a support to next step work. These concepts include: complexity, edge of chaos, self-organization, emergence, adaptation and evolution etc. Furthermore, with some applications of these notions and their quantifications in complex adaptive system, some inspirations of them for quatitative simulation and analysis of compelx adaptive system researches are illustrated. We hope this overview and these inspirations can improve the quantitative researches on simulation and analysis of complex adaptive system.Secondly, the speicification of complex adaptive system using architecture description language C2SADEL is introduced, and the architecture and formalized description is proposed. According to different architecture types, the specifications are discussed respectively, the study and involvement of Agent is discussed subsequently. A mathematical model of complex adaptive system is proposed, which is including its structure and dynamics. This model is constituted of definitions and specifications of all the entities.Thirdly, the quantitative model of ability of adaptive Agent is proposed, in order to solve the problem of reasoning the abilities of interactive agents in complex adaptive system. Then, several algorithms are advanced to schedule the tasks, including in static environment and in dynamic environment, whose efficiency is testified through several experiment results.Fourthly, a sign-based model of Stigmergy is proposed which picks up"Sign"as a key notion to understand it. A definition of"Sign"is given, which reveals the Sign's nature and exploit the significations and relationships carried by the"Sign". Then, a Sign-based model of Stigmergy is consequently developed, which captures the essential characteristics of Stigmergy. The basic architecture of Stigmergy as well as its constituents are presented and then discussed. The syntax and operational semantics of Stigmergy configurations are given. We illustrate the methodology of analyzing emergence in complex adaptive system by using our model.Fifthly, trust management is a fundamental concern in multi-agent system, especially in distributed multi-agent systems. Forthmore, trust mechanism will lead to emergence phenomenon, which is different to distinguish and analyse. So, a formalized descript method of certain problem in distributed trust model is proposed, and a simulated analysing method of distributed trust model based on agent is reached. This method is focusing on the devising of trust mechanisms. Starting with the relatively clear behavior of complex adaptive system micro level (i.e. agent's behavior, architecture, action rules), considering sufficient corresponding macro constrains, and building up a runnable simulation system, this approach studies the macro problems and the micro-macro relations of distributed trust model. Accordingly, in this thesis, at the micro level, the design scheme of trust mechanisms is analyzed and summarized; and based on this, a framework for implementing simulation agent is proposed. At the macro level, several system issues that must be taken into account in simulation-based analysis are discussed, including macro constrains, threat model, evaluating metric, and issues in simulation running. To demonstrate the effectiveness of this model, local trust model is analyzed based on it. The analysis also reveals the unpredictability and complexity of trust model.Finally, a case study of the battle simulation in military system is introduced, which is based on the thought and theory of complex adptive system.