基于城市风热环境的空间布局自动寻优方法研究

发布时间：2018-03-05 10:29

本文选题：城市风热环境　切入点：空间形态　出处：《深圳大学》2017年硕士论文　论文类型：学位论文

【摘要】：中国目前面临前所未有的高速城市化进程。城市建成区域建筑愈发密集的现状带来了一系列城市问题,城市微气候恶化也是突出问题之一。对城市规划、建设行业而言,有必要借助多学科的最新技术成果,提出缓解城市环境压力、提高城市生态和健康环境品质的有效方法,保证城市人口健康舒适生活的权益。本研究以此为出发点,以城市风热环境改善为具体目标,结合建筑学、城市规划学、计算机技术等多领域技术手段,试图建立一套城市设计自动生成与优化的科学方法,以适用于城市规划领域中从中观尺度的分区规划到较为微观的街区尺度城市设计的不同层面工作需要。在中观尺度的设计寻优方法研究中,采用城市气候地图的理论作为寻优的核心机理。根据数据表达的需要对区域进行栅格化处理,并设置与格栅对应的若干种单元模块,通过控制模块的建筑密度、高度等参数使之表征不同的城市建成区情况。以城市热累积最低为目标,在Rhino Grasshopper的软件平台上使用遗传算法进行各种模块不同比例、不同分布的方案自动寻优,以生成既定容积率条件下的环境最优布局。在微观尺度的设计寻优方法研究中,本研究首先进行了街区形态特征指标要素对街区整体风环境影响的机理研究。使用CFD软件工具建立半抽象模型进行模拟分析,提取代表不同模型形态特征的建筑密度、高度比、间隙率等要素,利用数理统计对形态要素的对街区整体风环境的影响程度进行回归分析,筛选相关要素并形成定量的回归方程。在此基础上再次利用Rhino Grasshopper软件平台,对该尺度下的建筑形态及布局进行生成方法研究。将中观尺度的设计生成的结果数据作为生成条件,代入微观尺度设计生成中,以上阶段得出的风速与形态要素指标间回归方程为寻优机理驱动遗传计算,经过代际遗传、进化获得该尺度的风环境最优的布局设计方案。作为全流程设计方法的验证和应用,选取深圳蛇口四海公园片区的实际情况进行生成设计的案例研究。结果显示,软件工作以及遗传算法对两个尺度的生成具有良好的兼容性,可在实际案例的具体约束条件下稳定进化方向,生成具有针对性的较理想的布局方案。验证了本研究提出的设计寻优方法及其相关系统的合理性、稳定性和有效性。
[Abstract]:China is now facing an unprecedented process of rapid urbanization. The increasing density of urban built-up areas has brought about a series of urban problems, and the deterioration of urban microclimate is one of the outstanding problems. It is necessary to put forward effective methods to alleviate urban environmental pressure, improve the quality of urban ecology and healthy environment, and guarantee the rights and interests of healthy and comfortable life of urban population. Taking the improvement of the urban wind-heat environment as the specific goal and combining with the multi-domain technical means such as architecture, urban planning and computer technology, this paper attempts to establish a set of scientific methods for automatic generation and optimization of urban design. In order to meet the needs of different levels of urban planning in the field of urban planning, from the mesoscale zoning planning to the more microscopic block scale urban design, in the study of the mesoscale design optimization method, The theory of urban climate map is adopted as the core mechanism of optimization. According to the need of data expression, the area is rasterized, and several kinds of unit modules corresponding to the grid are set up to control the building density of the module. Taking the lowest urban thermal accumulation as the goal, genetic algorithm is used to automatically optimize various modules in different proportion and distribution on the software platform of Rhino Grasshopper. In order to generate the optimal layout of the environment under the condition of given volume ratio, in the study of the design optimization method at the micro scale, In this study, the mechanism of the influence of the index elements of the block shape characteristics on the overall wind environment of the block is studied. The semi-abstract model is established by using the CFD software tool to simulate and analyze, and the building density and height ratio representing the morphological characteristics of the different models are extracted. Based on the regression analysis of the influence degree of morphological factors on the overall wind environment of the block by mathematical statistics and other factors, the relevant factors are screened and a quantitative regression equation is formed. On the basis of this, the Rhino Grasshopper software platform is used again. The generation method of building form and layout under this scale is studied. The result data generated by the mesoscale design is taken as the generating condition, and then the micro-scale design is added to the generation. The regression equation between wind speed and morphological factors obtained in the above stage is the optimization mechanism driving genetic calculation. Through intergenerational heredity, the optimal layout design scheme of wind environment of this scale is obtained by evolution, which is used as the verification and application of the whole process design method. A case study on the generation design of Shikou Sihai Park in Shenzhen is carried out. The results show that the software and genetic algorithm have good compatibility with the generation of the two scales. The evolutionary direction can be stabilized under the specific constraints of practical cases, and a more targeted and ideal layout scheme can be generated. The rationality, stability and effectiveness of the proposed design optimization method and its related systems are verified.
【学位授予单位】：深圳大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU119

【相似文献】