Abstract: The intensification of global temperature changes has caused frequent urban natural disaster events related to climate, seriously affecting the public safety of cities. At present, the construction of the urban waterlogging disaster resilience evaluation system is mostly from the perspective of various components of urban systems, and the subjective evaluation methods are used in the evaluation model construction. Based on this, 27 specific indicators were selected to establish resilience evaluation system for urban waterlogging disaster from 3 aspects such as resistance, recovery capability and adaptive capacity, and the urban waterlogging disaster resilience evaluation model based on the projection pursuit model was established. The real code accelerated genetic algorithm was adopted to optimize the solution. For high-dimensional data, the proposed model could directly identify its structure and features from the data itself, avoiding the limitations of subjective judgment and making the evaluation results more authentic and reliable. Finally, the proposed model was used to evaluate and analyze the resilience of urban waterlogging disasters in 18 prefecture level cities in Henan Province. The research results show that there are obvious differences in the resilience levels of urban waterlogging disasters in 18 prefecture level cities in Henan Province, among which Zhengzhou has the highest resilience evaluation value, followed by Luoyang and Jiaozuo, and Puyang has the lowest resilience evaluation value. The medical conditions, economic level and infrastructure construction have a great impact on the resilience level of urban waterlogging disasters.

Key words: resource and environment; waterlogging disaster; urban resilience; projection pursuit; RAGA

摘要： 全球气温变化导致与气候相关的城市自然灾害事件频发，严重影响到城市公共安全。目前针对城市内涝灾害韧性评价体系的构建多是从城市系统各个组成结构角度出发，且在评价模型构建方面均采用了主观性较强的评价方法。笔者从抵抗力、恢复力和适应力3个方面选取27个具体指标建立了城市内涝灾害韧性评价指标体系，基于投影寻踪模型构建了内涝灾害韧性评价模型，采用实码加速遗传算法进行优化求解；针对高维数据，此模型能直接从数据本身找出其结构和特征，避免了主观判断的局限性，使评价结果更加真实可靠；最后利用该模型对河南省18个地级市的城市内涝灾害韧性进行评估分析。研究结果表明：河南省18个地级市的城市内涝灾害韧性水平存在明显差异，郑州市的韧性评价值最高，其次是洛阳市和焦作市，濮阳市的韧性评价值最低；医疗条件、经济水平和基础设施建设对城市内涝灾害韧性水平影响较大。

关键词: 资源环境；内涝灾害；城市韧性；投影寻踪；实码加速遗传算法