基于神经网络综合建模的区域城市群发展脆弱性评价——以滇中城市群为例

城市脆弱性是评价城市发展韧性状况的有效测度。目前我国对于城市脆弱性的研究以运用统计学方法评价特殊城市为主,尚未形成对区域城市群具有普遍适用性的科学客观评价方法。以滇中城市群为例,从环境系统、经济系统、社会系统三个方面综合构建城市脆弱性评价体系,采用熵值法和BP神经网络综合建模方法,对2007—2016年10年间滇中城市群的城市脆弱性进行评价。结果表明:滇中城市群的城市脆弱性总体呈现下降趋势,但城市组团之间差异较大,呈现出发展中的不均衡性。评价结果对滇中城市群韧性发展规划具有重要指导意义,为区域城市群发展脆弱性研究提供一种科学评价方法。

Vulnerability assessment of urban agglomeration based on neural network model: A case study of Central Yunnan Urban Agglomeration

Urban vulnerability is an effective measure to evaluate urban development resilience. At present, most studies on urban vulnerability focus on the special city and use the statistical method in China, and there is no urban vulnerability assessment method that is applicable to all the regional urban agglomerations. Taking the Central Yunnan Urban Agglomeration as an example, this paper develops an assessment system of urban vulnerability from three aspects of environmental system, economic system and social system. Entropy method and back propagation neural network model are used to evaluate the urban vulnerability of the study area from 2007 to 2016. The evaluation shows that the vulnerability of this urban agglomeration has a declining trend on the whole, but there is a big difference between the urban groups, which shows an unbalanced development. The evaluation results have reference significance for the planning and resilience development of the Central Yunnan Urban Agglomeration, and provide a scientific evaluation method for the study on the comprehensive development vulnerability of urban agglomeration.