关键词: 多智能体系统, 智能体, 复杂系统, 城市用地布局, 由下至上, NetLogo

Abstract:

Since the human society has entered the 21st century, its rapid developing status not only provides the necessary impetus for both technology advancement and economic growth, but also poses a challenge on the sustainable development of ecological environment and social culture in the city. As we can see, the current quantity of construction land demanded by the city has already exceeded the rational carrying capacity of natural environment, and the incongruity between them will cause many kinds of problems related to the distribution of land resources and environment protection. Therefore, how to ensure the city can get the long-range and healthy development while fulfilling the increasing demand of construction land is becoming a major concern among both government administrators and urban designers. Although setting up a simulation model for the research area is one of the valid ways to evaluate the possible outcomes of future development, but as different components of urban system would interact with each other intricately and uncontrollably, and it will bring non-linear and dynamic impacts on the spatial development of urban land use, the assumption of traditional modelling methods about the static and equilibrium status of simulation structure will not fully match with the reality and may cause some significant deviations during its simulation process. So it is necessary for us to find a new way to better describe the dynamic characteristics of real world in the simulation model. Under this situation, the emergence of complex system theory can provide researchers an useful alternative to model the complex and mutual interactions among different individual components, and its “bottom-up” simulation structure is highly consistent with the mechanisms that urban system is developed from development activities in micro-level. Therefore, with the help of complex system theory, people can change their traditional static analysis thinking into a more active pattern and predict the future development trend for urban land use more accurately. As a result, in order to effectively solve the problems between arranging urban land use development and protecting the ecosystem, this research decided to use the multi-agent system as a simulation tool to build the land use development model based on the knowledge of complex system theory, urban economics and geography information system. Unlike traditional modelling methods, the main research mechanism of MAS (multi-agent simulation) model is to make use of the interaction behaviors of different kinds of developer agents to predict the future development trend for the construction areas in San Diego, following the simulation structure of bottom-up. Also, this research has proved its simulation accuracy through both point-by-point comparison and geometrical evaluation. After posting a general summary about the key points of the entire research, this paper analyzed the existing problems of the current land use simulation model, and proposed several possible ways to improve its effectiveness in the future.

Key words: multi-agent system, agent, complex system, urban land use, bottom-up, NetLogo