Design and Control of Self-organizing Systems--论文代写范文精选

一般介绍复杂思维,由于设计自组织系统需要一个非经典的思想,而实际提出了复杂性和自组织的概念。为了说明方法,以三个案例研究。自组织的交通灯控制器和多主体模拟系统研究,这优于传统方法。下面的paper代写范文将进行详述。

ABSTRACT
Complex systems are usually difficult to design and control. There are several particular methods for coping with complexity, but there is no general approach to build complex systems. In this thesis I propose a methodology to aid engineers in the design and control of complex systems. This is based on the description of systems as self-organizing. Starting from the agent metaphor, the methodology proposes a conceptual framework and a series of steps to follow to find proper mechanisms that will promote elements to find solutions by actively interacting among themselves. The main premise of the methodology claims that reducing the “friction” of interactions between elements of a system will result in a higher “satisfaction” of the system, i.e. better performance. 

A general introduction to complex thinking is given, since designing self-organizing systems requires a non-classical thought, while practical notions of complexity and self-organization are put forward. To illustrate the methodology, I present three case studies. Self-organizing traffic light controllers are proposed and studied with multi-agent simulations, outperforming traditional methods. Methods for improving communication within self-organizing bureaucracies are advanced, introducing a simple computational model to illustrate the benefits of self-organization. In the last case study, requirements for self-organizing artifacts in an ambient intelligence scenario are discussed. Philosophical implications of the conceptual framework are also put forward.

INTRODUCTION
Our world becomes more complex every day. To cope with it, the systems we design and control also need to become more complex, increasing the complexity of the world. This increase in complexity is characterized by a growth in number and diversity of elements of systems and their interactions. Every year there are more people, more computers, more devices, more cars, more medicines, more regulations, more problems. Since complexity has crawled into all the aspects of our lives, its study is very important for all disciplines. Traditional approaches are becoming obsolete, as they were developed for a simpler world. Thus any advancement in the general understanding of complex systems (Bar-Yam, 1997) will have a potential impact in sciences, engineering, and philosophy. “Classical” approaches are still very useful, but only in problem domains that are static and comprehensible. An exact solution can be found, and this solution will hold. But with an increase of complexity, problem domains become dynamic, requiring for dynamic solutions that will be able to adapt to the changes in the problem domain (Ashby, 1947a).

Classical Thinking 
The majority of scientific models—as well as much of our intuitive understanding—implicitly rely on a “classical” or Cartesian mode of thinking, which is expressed most explicitly in the classical or Newtonian mechanics that dominated the scientific worldview until the beginning of the 20th century. It is based on the following assumptions (Heylighen, 1990a): 
• reductionism or analysis: to fully understand a system you should decompose it into its constituent elements and their fundamental properties. 
• determinism: every change can be represented as a trajectory of the system through (state) space, i.e., a sequence of states, following fixed laws of nature. These laws completely determine the trajectory towards the future (predictability) as well as towards the past (reversibility). 
• dualism: the ultimate constituents of any system are particles, i.e., structureless pieces of matter (materialism). Since matter is already completely determined by mechanistic laws, leaving no freedom for intervention or interpretation, the only way we can include human agency in the theory is by introducing the independent category of mind. Complexity 11 
• correspondence theory of knowledge: through observation, an agent can in principle gather complete knowledge about any system, creating an internal representation whose components correspond to the components of the external system. This establishes a single, true, objective mapping from the realm of matter (the system) to the realm of mind (the representation). 
• rationality: given such complete knowledge, in its interaction with the system, an agent will always choose the option that maximizes its utility function. Thus, the actions of mind become as determined or predictable as the movements of matter.

These different assumptions are summarized by the principle of distinction conservation (Heylighen, 1989, 1990b): classical science begins by making as precise as possible distinctions between the different components, properties and states of the system under observation. These distinctions are assumed to be absolute and objective, i.e., the same for all observers. They follow the principles of Aristotelian logic: a phenomenon belongs either to category A, or to not A. It cannot be both, neither, in between, or “it depends”2 . 

The evolution of the system conserves all of these distinctions, as distinct initial states are necessarily mapped onto distinct subsequent states, and vice versa (causality, see Heylighen (1989)). Knowledge is nothing more than another such distinction-conserving mapping from object to subject, while action is a mapping back from subject to object. Certainly, we know that these assumptions represent ideal cases that are never realized in practice. Yet, most educated people still tend to assume that a complete and deterministic theory is an ideal worth striving for, and that the scientific method will lead us inexorably to an ever closer approximation of such objective knowledge. However, the lessons from complexity research point in a different direction (Morin, 2006).

Objectivity or Subjectivity? 
Contextuality! As we saw in Chapters 2 and 3, the observer is essential in determining features of a system such as complexity, self-organization, cognition, intelligence, life, and consciousness. Does this mean that we are doomed to subjectivity? No, since all of these properties are applied to an external abs-being, i.e. independent of the observer, and can be scientifically contrasted (Popper, 2002). Like this, we can hope to find pragmatically when it is more useful to attribute certain properties to certain systems, but always within a certain context (Gershenson, 2002c; Edmonds, 2001). This is because the “usefulness” of describing an abs-being with a particular rel-being may change across contexts. This idea was already proposed, among others, by constructivism (von Glasersfeld, 1984; Riegler, 2005) and second-order cybernetics (Heylighen and Joslyn, 2001), but seems to be lacking in most sciences still, as researchers keep on trying to find a purely objective and absolutely true nature of reality. We can only approach reality as observers, so we cannot ignore our inherent subjectivity. But a feedback between reasoning and experience will be able to help us discern which models are more appropriate for different circumstances. 

The conflict between objectivity and subjectivity goes back to the debate between rationalism and empiricism, which can be traced to the opposition between the teachings of Parmenides and Heraclitus. But actually, when they spoke about the being, the former referred to the abs-being (static, unique, absolute), while the latter to rel-beings (dynamic, multiple, relative). Which one was right? Neither and both, since they each refer to different rel-beings appropriate for different contexts. What we gain with contextuality is the ability to switch between approaches as different contexts demand, since we understand that in practice a “true” model or description of the world cannot be reached. This has implications not only for science and philosophy, but also for society in general. Contextuality gives us the possibility of avoiding social friction by Tolerance, Courtesy, and Compromise. With a purely “ob- Conclusions 143 jective” worldview (Aerts et al., 1994), one cannot accept that more than one “truth” may exist, leading to fanaticism and orthodoxy, e.g. Nazism, terrorism. A purely “subjective” worldview would not be able to discern which ideas are useful for society, since it gives equal value to any of them, e.g. certain non-academic postmodern worldviews. However, a “contextual” worldview is able to find rel-beings valid for specific contexts (contrasted with experience (Popper, 2002)) and to accept different rel-beings in different contexts. Like this, it not only tolerates, but even interacts with them1 . 

Only a contextual worldview will be able to reduce friction and promote synergy to increase social satisfaction. 8.3.2 The Benefits of Self-organization “We can only see a short distance ahead, but we can see plenty there that needs to be done” —Alan M. Turing The case studies presented in this thesis tried to show the benefits of modeling complex systems as self-organizing. The self-organizing traffic lights (Chapter 5) are able to coordinate dynamically according to actual traffic densities, adapting effectively to changes in the traffic densities and flows. Moreover, they are robust and distributed, having several benefits above traditional “blind” methods apart from considerably reducing travel times. The ideas and simulations presented for self-organizing bureaucracies (Chapter 6) showed that using the Methodology in organizations may enable them to improve constantly their performance, adapting to changing demands. Finally, the ideas presented to achieve the self-organization of artifacts (Chapter 7) argued that it is possible to design adaptive technologies that will be able to cope with changes of specifications, which will certainly occur in our information-centered world. This would enable devices to learn by themselves new meanings and ways of interaction, potentially producing novel functionalities as they coordinate. 

CONCLUCION
As we can see, one of the main benefits of an engineered selforganizing system is that of adaptation. This would be redundant for a static problem domain. However, most complex systems have dynamic problem domains, where solutions change constantly. A self-organizing system will be able to seek by itself new solutions, having more potential and robustness than a traditional approach. 1Paradoxes are not an impediment for this (Gershenson, 1998b, 1999). 144 Philosophical Implications Any system is liable to make mistakes (and will make them in an unpredictable environment). But a good system will learn from its mistakes. This is the basis for adaptation. It is pointless to attempt to build a “perfect” system, since it is not possible to predict future interactions with its environment. What should be done is to build systems that can adapt to their unexpected future and are robust enough not to be destroyed in the attempt. Self-organization provides one way to achieve this, but there is still much to be done to harness its full potential.(论文代写)

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。(论文代写)
更多论文代写范文欢迎访问我们主页 www.51due.com 当然有论文代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X(论文代写)