Exploration into self-organised criticality of urban spatial systems evolution - applicaion of the bounded rationality paradigm

• Main Authors: Kao Hong-shwen, 高宏軒
• Other Authors: Lai Shih-kung
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/56403115975356989496

碩士 === 國立中興大學 === 都市計劃研究所 === 86 === There is a recent shift from the top down to the bottom top approach to explor ing into urban spatial models partly because of variety and complexity of urba n development. The top down approach assumes that there exists an equilibrium point in urban development, and econometric and human ecological models for ur ban development are constructed to search for that equilibrium point. The bott om up approach starts from individual agents'' preferences and behaviors to exp lore into the interaction between these agent and the external environment. Th epurpose is to understand possible links between individual agents'' behavior a nd the urban evolution as a whole. Because the number of individual agents can be large and the choices combination of these agents can thus be complex, view ing a city as a complex spatial system becomes a new approach to understanding urban development evolution. Land use change in urban space is affected not only by external factors, but a lso by the interaction among land uses nearby which forms one of the important factors determining the land use model. Land use models based on mathematical constructs usually assume that market information is perfect and that spatial decision makers can attain full information, ignoring the differences in indiv idual agents'' ability to acquire information and make decisions. Therefore, de velopment decision makers'' behavior is subjec to bounded rationality. Given th is assumption, the research applies two dimensional cellular automata to compu tersimulations to investigate into how land wus changes and observe what patte rnsof large scale land use would emerge from the interaction among individual agents. In the experimental design, the research views each parcel of land in an artificial city as a cell whose land use pattern transition is selected fro m aset of development decisions contingent on the land use pattern in its neig hborhood.Development decision rules are constructed through a questionnaire su rvey as the development database in the simulation experiments. The experiment al design is based on the notion that a city is a complex spatial system and t hrough the bottom up approach to spatial evolution, we can conduct experiments of computer simulation to explore into the complex evolution phenomena of the system. The experimental results show that the scales of clusters of cells with the sa ame land use and the associated frequencies tend to self-organize into a power law. According to Wolfram''s fourth classifications of cellular automata, the f ourth category of transition rules will yield self-organized, complex structur es. Through a questionnaire survey to construct evolution rules of bounded rat ionality and using the bottom top interaction approach as in the original cell ular automata, the system evolves toward fractal spatial distribution as an in dication of self-organized criticality. It can be inferred that behavior of bo unded rationality is one of the factors that leads to complex spatial structur es implied by Wolfram*s fourth category of rules in cellular automata. The exp erimental results are similar to those obtained from real urban spatial datain that fractal phenomena are common in empirical studies. It follows that land d evelopment of bounded rationality is perhaps the basis of individual agents'' b ehavior of more realistic spatial interaction models.