【作者】 申晓明；

【导师】 谢慧才；

【作者基本信息】 汕头大学 ， 结构工程， 2003， 硕士

【摘要】 本文的工作是国家自然科学基金重点项目(59938180)“基于GIS的城市综合减灾评估决策系统”的一部分。 风灾是自然灾害中影响非常广泛的一种，其中台风灾害又是风灾中破坏力最大、造成损失最严重的一种。风灾易损结构物(WDS)是在台风灾害中最易受到破坏的结构物，但目前国内外对这类结构物的研究尚很少见到。 本文提出了一种全新的研究方法，对这类结构进行了损伤评估，并实际用于汕头市防风研究，建立了汕头市风灾易损结构物的损伤评估系统，取得了很好的效果。 本文以户外广告牌为例，在大量调研的基础上，建立了典型结构的力学模型，然后采用功能强大的Ansys有限元软件建立有限元模型。鉴于结构风灾损伤的特点，进行了非线性(包括大变形和弹塑性)动力学分析，并提出了损伤评估标准。在此基础上，提出了神经网络方法进行易损结构物的损伤评估。利用有限元分析，进行大量的样本计算，为神经网络方法提供了样本学习库，建立了结构物损伤与致灾因素的关系。采用神经网络的评估方法，避免了对结构单体进行逐个繁杂的有限元分析，大大提高了效率，并为进一步建立自动化的评估系统奠定了基础。 为了使本评估方法能用于实际，本文经大量调研，建立了基于地理信息系统(GIS)的汕头市风灾易损结构物空间数据库。在基础地理信息系统基础上，调用风灾易损结构物损伤评估方法，建立了汕头市风灾易损结构物损伤评估系统。该系统可模拟台风侵袭汕头时，该类结构物的损伤状况，并在地图上显示位置分布情况，可起到预警、减灾和灾害评估的作用。同时本文所提出的方法可推广应用于其他类型结构的减灾防灾研究中。更多还原

【Abstract】 The job of this thesis is part of the project named GIS-BASED EVALUATION AND DECISION SUPPORT SYSTEM FOR URBAN HAZARD RESISTANCE (59938180), which is sponsored by National Natural Science Foundation of China.Wind hazard is the one which has the widespread influence in the nature disaster, and hurricane hazard is the most devastative and disastrous one. The most destructible structure in the hurricane is the Wind Destructible Structure(WDS) which lacks study in the world.This thesis puts a kind of brand-new approach forward the research of WDS damage assessment and applies it in the study of mitigation of wind-excited hazard in Shantou. The application obtains satisfactory results.This thesis takes the outdoor bill-board as an example, on the basis of a large number of survey and study, sets up the mechanics model of typical WDS and the finite element model adopting Ansys, a kind of powerful finite element software. Seeing that structure characteristic of damage in the wind hazard, non-linear dynamics analysis was carried on, and has proposed the standard of damaging assessment. On this basis, neural network was used to assess the damage of WDS. A large number of samples calculate have been done utilizing FEA analyse. This offers learning samples to neural network method. The relation between the structure damage and the factors of causing disaster was set up. Adopting neural network method avoids carrying on complex analyse one by one to structure. It enhances the efficiency greatly and establishes the foundation for the automative assessment system further.For making the assessment method be used in reality, this thesis set up the Shantou’s WDS spatial database after a large amount of survey. Based on basic GIS, applying the standard of WDS damaging assessment, the WDS damage assessment system of Shantou city has been built. The system can simulate the state of WDS’s damage when hurricane lands on Shantou city and show the distributing complexion, and forecast,decrease,assess damage. The method that this thesis bring forward can be extended to the research of mitigation of wind-excited hazard for other type ofstructure.更多还原