滨海城市自然灾害风险评估与控制方法的基础研究

发布时间：2018-09-09 12:05

【摘要】：如今全球遭受到的自然灾害的强度和频率不断增加，干旱、洪水、飓风、地震和海啸等灾害给人类社会带来的损失不断加剧。城市则是受自然灾害影响最大的区域，而在所有城市中滨海城市更是人口、财富、技术聚集的核心区域，是自然灾害易发和频发的区域。全方面地对滨海城市自然灾害风险评估与风险控制方法进行研究显得尤为必要。因此，本文以滨海城市自然灾害为研究对象展开了灾害风险评估与风险控制方法的基础性研究，主要研究内容包括了如下几个方面：（1）对滨海城市自然灾害等级界定方法进行了研究，提出了基于模式识别理论的多指标滨海城市自然灾害等级界定算法，它是一种多指标下的模式分类方法。通过研究发现基于多指标的滨海城市自然灾害等级的界定实际上可以简化成为一个多输入单输出的模式判别问题。因此，本文建立了线性模式判别函数对滨海城市灾害等级进行划分。在线性判别函数系数的确定上，结合增量固定算法编制了相关的计算机程序模块，以滨海城市常见的风暴潮灾害数据为例，实现了判别函数系数的快速计算，达到对滨海城市自然灾害级别的快速界定，说明了该算法的可用性。（2）对滨海城市自然灾害风险预测的方法进行了研究。本文通过研究发现自然灾害本身具有混沌的特征，而混沌理论能够很好地解释自然灾害所表现出的混沌现象，同时对灾害未来的发展规律进行预测。因此，本文提出了基于混沌理论的自然灾害预测模型。通过研究表明，混沌理论在自然灾害的预测方面是可行的，可以取得较为理想的结果。与此同时，在自然灾害风险等级预测方面，本文提出了基于神经网络的多指标滨海城市自然灾害风险等级预测模型。以风暴潮灾害为例，建立了三层BP神经网络的自然灾害致灾等级预测模型，编制了分析程序并进行了预测计算，通过实例分析证明在通过主成因识别法对风暴潮致灾的主要因素进行识别的基础上，神经元网络模型能够较为精确地预测出自然灾害的灾害等级，预测效果较好。（3）对滨海城市自然灾害风险的定量化评估方法进行了研究。本文以风暴潮灾害与地震灾害为例进行了灾害风险定量化评估方法研究，构建了定量化评估模型。在风暴潮灾害风险的定量化评估模型中，讨论了单一承灾体和行业承灾体的损失率计算模型，并通过潮灾损失定量统计模型对滨海城市的潮灾损失进行了汇总统计。在地震灾害风险的定量化评估模型中，提出了基于烈度的单体建筑物损失评估的震害影响因子法和基于历史灾情数据统计的群体建筑物损失评估方法，最后按照震害损失定量统计模型对震害损失进行了汇总统计。（4）对滨海城市自然灾害风险空间可视化评估方法进行了研究。本文运用GIS技术针对特定灾害进行了可视化的模拟，并在对研究区域进行了评价单元划分的基础上评估了灾害的损失，同样以风暴潮灾害与地震灾害为例进行了灾害风险可视化评估方法的研究。在风暴潮灾害风险的可视化评估过程中，本文利用GIS中的网格计算模型模拟了风暴潮的影响范围与范围内的水深分布情况，结合潮灾定量化评估模型，编程实现了风暴潮灾害损失评估模块，并进行了潮灾风险可视化评估的实例分析。在地震灾害风险的可视化评估中，本文查阅了研究区的地震烈度衰减模型利用GIS技术绘制出了地震的影响范围，并根据震害损失定量统计模型对研究区的人员伤亡、经济损失、建筑物破坏情况等进行了评估，编制计算机程序模拟出了地震灾害对研究区造成的损失。该方法全面可视化地表现了灾区的受灾情况，达到了较好的效果。（5）以青岛市为例进行了滨海城市自然灾害的可视化模拟和风险评估。以青岛市作为研究区，通过对青岛市历史上灾害致灾因子强度的概率分析得出了区域未来可能遭受到的灾害的强度。根据定量化和可视化评估方法评估了青岛市未来一旦遭受到风暴潮和地震灾害时的损失情况，运用GIS技术对灾害影响区域进行了模拟，统计出了青岛市各个社区灾害的损失情况，并在地图上标示出了灾害的风险区，生成了城市中各类承灾体的破坏专题图，评估效果较为满意。（6）对滨海城市自然灾害风险控制方法进行了研究。本文把构建滨海城市自然灾害预警与应急信息系统作为滨海城市自然灾害风险控制的主要手段，并在ArcGIS Server平台上讨论了系统的逻辑结构设计、功能设计与数据库设计，，并对系统的关键技术和核心功能的实现作了相关的研究，建立了滨海城市自然灾害风险控制信息系统，实践证明该系统在滨海城市自然灾害风险的评估与控制中能起到较好的辅助决策作用。
[Abstract]:Nowadays, the intensity and frequency of natural disasters are increasing all over the world. Droughts, floods, hurricanes, earthquakes and tsunamis are causing increasing losses to human society. Cities are the areas most affected by natural disasters. In all cities, coastal cities are the core areas of population, wealth and technology accumulation, and natural disasters. It is necessary to study the risk assessment and risk control methods of natural disasters in coastal cities in all aspects. Therefore, the basic research on risk assessment and risk control methods of natural disasters in coastal cities is carried out in this paper. The main research contents include the following aspects :
(1) The method of natural disaster classification of coastal cities is studied, and a multi-index algorithm of natural disaster classification based on pattern recognition theory is proposed, which is a multi-index pattern classification method. This paper establishes a linear pattern discriminant function to classify coastal city disasters. In the determination of linear discriminant function coefficients, a computer program module is compiled with the incremental fixed algorithm, and the storm surge disaster data of coastal cities are taken as an example. The fast calculation of discriminant function coefficients and the fast definition of natural disaster level of coastal cities show the feasibility of the algorithm.
(2) The risk prediction methods of natural disasters in coastal cities are studied in this paper. It is found that natural disasters have chaotic characteristics, and chaos theory can well explain the chaotic phenomena of natural disasters and predict the future development of natural disasters. The research shows that chaos theory is feasible in the prediction of natural disasters and can obtain ideal results. At the same time, in the prediction of natural disaster risk grade, a multi-index forecasting model of coastal city natural disaster risk grade based on neural network is proposed. As an example, a three-layer BP neural network model for predicting natural disasters is established, the analysis program is compiled and the prediction calculation is carried out. It is proved that the neural network model can accurately predict natural disasters on the basis of identifying the main factors causing disasters by storm surge through the method of principal cause identification. The prediction of disaster grade is better.
(3) The quantitative assessment method of natural disaster risk in coastal cities is studied. Taking storm surge disaster and earthquake disaster as an example, the quantitative assessment method of disaster risk is studied, and the quantitative assessment model is constructed. In the quantitative assessment model of earthquake disaster risk, the seismic damage impact factor method based on intensity and the group building loss evaluation based on historical disaster data statistics are proposed. Finally, according to the statistical model of earthquake damage loss, the losses of earthquake damage are summarized.
(4) The spatial visualization assessment method of natural disaster risk in coastal cities is studied. In this paper, GIS technology is used to visualize the simulation of specific disasters, and the loss of disasters is evaluated on the basis of the division of evaluation units in the study area. In the process of visualization assessment of storm surge disaster risk, this paper simulates the distribution of water depth in the scope and scope of storm surge by using grid computing model in GIS. Combining with the quantitative assessment model of tide disaster, the loss assessment module of storm surge disaster is programmed and realized, and the risk of tide disaster can be calculated. In the visual evaluation of earthquake disaster risk, this paper consults the seismic intensity attenuation model of the study area and draws out the influence range of the earthquake by using GIS technology, and evaluates the casualties, economic losses and building damage in the study area according to the quantitative statistical model of earthquake damage loss. The computer program simulates the damage caused by earthquake disaster in the study area. The method shows the disaster situation of the disaster area visually and comprehensively, and achieves good results.
(5) Taking Qingdao as an example, the visual simulation and risk assessment of natural disasters in coastal cities are carried out. Taking Qingdao as a research area, the intensity of disasters that may occur in the future in Qingdao city is obtained through the probability analysis of the intensity of disaster-causing factors in the history of Qingdao city. In the last few years, the loss of storm surge and earthquake disaster was simulated by using GIS technology. The loss of disasters in different communities in Qingdao was counted. The risk area of disasters was shown on the map, and the damage thematic maps of various disaster-bearing bodies in the city were generated. The evaluation results were satisfactory.
(6) The risk control methods of natural disasters in coastal cities are studied. In this paper, the early warning and emergency information system of natural disasters in coastal cities is constructed as the main means of risk control of natural disasters in coastal cities. The logical structure design, function design and database design of the system are discussed on the platform of ArcGIS Server. The key technology and the realization of the core function are studied, and the coastal city natural disaster risk control information system is established. The practice proves that the system can play a better role in the coastal city natural disaster risk assessment and control.
【学位授予单位】：中国海洋大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TU984.116;X43

【参考文献】