Cellular Automata for Studying Historical Spatial Process of Traditional Settlements Based on Gaussian Mixture Model: A Case Study of Qiaoxiang Village in Southern China

摘要

The spatial evolution of traditional rural settlements is a natural process that occurs without professional development planning. With rural spatial organization now being stimulated by China's economic structure adjustment, making top-down plans match bottom-up spatial principles is important for the continuity of rural communities. In this research, a Cellular automaton based model was introduced to describe the historical spatial rules of rural settlements based on a traditional village in Southern China. In order to reproduce the rural spatial characteristics and laws that were close to the actual situation truthfully and distinctly, in this model, we refined the growth process into three cyclical steps, and used a Multivariate Gaussian Mixture Model for the evaluation of spatial transition potential. Through historical simulation, evaluation of the current status, and future spatial predictions, results showed that: (1) the growth during the whole spatial process maintained logical consistency for optimal selection; (2) certain houses in the southeast region should be carefully protected as a core living culture exhibit, while the best utilization of other homes should be considered; and (3) the most probable region of growth for tourist services or new housing estates is scattered along the river in the central area of the village.