Architecture and Design--论文代写范文精选

在这篇paper代写范文中,首先概述了架构的需求,这些需求将指导我们。接下来,详细说明建筑设计的完整系统,即应用架构模式。

Introduction
  We have given the foundations for the design of a new cognitive agent-based computational social simulation model. In this chapter, we will explain the design of a new cognitive architecture RBot, an objectoriented architecture that can easily be plugged-in into a multi-agent platform. In this chapter, we will design such a platform or task environment as well; our so-called Multi-RBot System (MRS). However, we will rst express again the need for such a new actor1 architecture. Hence, the rst question for the reader should be: why developing another multi-actor simulation model and why not using existing software toolkits that give the possibility to model actors? 

  Although the problem of nding multi-agent toolkits is no problem (there are many available2 ), many agent toolkits are developed with specic areas of research in mind. Such toolkits model for example social phenomena and are focused on the overall behaviour of populations (large populations), waiting queues and bottleneck studies. Or they focus on the intelligent agent itself and when that is the case, they focus only on certain aspects of the agent. Even when they focus on many aspects of the individual interacting with the environment, the assumptions of the model are most of the time not built upon ndings in cognitive psychology (such as ACT-R does). Therefore, an initiative was taken to develop a new cognitive architecture from scratch, of course with other existing agent/actor technologies in mind. 

  This chapter will follow a structure that starts with an overview of the requirements of the architecture and these requirements will guide the designprocess of our architecture. Next, a section will follow that elaborates the gen- 1 eral architectural design of the complete system, i.e. what architectural patterns are applied. Following up this section, we apply a bottom-up approach in designing the system. We will rst start a design process of a single actor and then go into details of how we will unite multiple actors into a multi-actor system. The chapter will end with a short discussion that functions as an introduction to the next chapter. That chapter will discuss demonstrative experiments and the conguration of the actors and the task environment in detail. 

  General Requirements and Overview of the Architecture In this section, we will rst give the general requirements that are necessary for our architecture. Next, we will give a short explanation of patterns and then choose some of those patterns for the general design of our architecture. The general requirements depend on the research questions we have posed in chapter 1. We will make use of the research questions (11.1 until 1.3) and the chapters about MAS, the social and the cognitive actor that deliver detailed requirements for the design of our RBot architecture. The research questions express three general requirements that the newly designed architecture should full: 
  1. Research question 1.1 indicates that the system should support social behaviour and that such behaviour should be explained in terms of individuals and their interrelations. Distributed systems that are applied in DAI are suitable candidates for modelling such a system. The latest development in the area of distributed systems is the Multi-Agent System (see chapter 2), which is an appropriate methodology for modelling entities that interact with each other and the environment. 
  2. Chapter 3 answered research question 1.2 and discussed what was necessary for an actor in a MAS to exhibit social behaviour. The discussion made clear that the social construct is a (shared) unit of knowledge that can guide social behaviour. The social construct as a representation in the mind of the actor can be expressed as a data structure in the mind of the actor. Therefore, an actor needs a mind where representations can nd their place.
  3. Research question 1.3 addresses this issue, and more specically, what kind of actor can handle representations, signs and social constructs. Chapter 4 explained the cognitive actor and stated that the ACT-R architecture can very well serve as an architecture for our cognitive plausible actor. Therefore, the requirement of our cognitive architecture is to inherit the model, the characteristics and the cognitive mechanisms of ACT-R in order to design a cognitive plausible actor. 

  The design of an architecture needs a methodology that supports proper design. Patterns provide solutions and methods for solving problems that often reoccur. In this chapter, we will adopt patterns and apply them for nding an appropriate design that sufces our requirements brought forward by our research questions. In the next section, we will rst explain shortly what patterns are and with help of those patterns, we will create an abstract design of our cognitive agent-based social simulation system.

 Overview of the architecture 
  The overview of the architecture has been designed based on a combination of a set of general architectural patterns. We will follow the design of a MAS (requirement 1), which requires that the components of the architecture should have a high degree of autonomy. The Component Interaction View (CIV) (Zdun & Avgeriou, 2005, p. 28) contains individual components that interact with each other but retain their autonomy, i.e. they merely exchange data but do not directly control each other. Interaction can be performed synchronously or asynchronously and can be message-based or through direct calls. These components can be distributed. Therefore, the CIV is closely connected to the Distributed Communication View (DCV). The CIV and the DCV are suitable to be applied for designing a MAS, because of their characteristics of autonomy and distribution, respectively. (paper代写)

  The overview3 of patterns associated with CIV and DCV is displayed in gure 5.1. In most multi-agent simulation systems, autonomous actors `live' in a task environment, which is a separate process that is inaccessible but can be inuenced by the actors. Hence, in the design of our architecture, we have made the decision to separate the task environment and the actors. Because the task environment (supplier of the environment) and the couplings with the actors (clients: consumers of the environment) are known at design time, the Explicit Invocation4 pattern is a general architectural pattern that can be applied for our architecture. Within the Explicit Invocation we can make a decision between the Client-Server (CS) pattern and the Peer-to-Peer (PtP) pattern. The CS pattern consists of a client and a server, two independent components that are not equal to each other: one, the client, is requesting a service from the other, the server. On the other hand, the PtP pattern is somewhat similar to the CS pattern, however all components are equal and can provide as well as consume services. The pattern applied in our design is the Client-Server pattern, i.e. the components client and server are not equal; the server is the task environment and the actors are the clients.(paper代写)

  51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。
  更多paper代写范文欢迎访问我们主页 www.51due.com 当然有paper代写需求可以和我们24小时在线客服QQ:800020041 联系交流。-X(paper代写)

标签：paper代写  留学生作业代写  assignment代写