【作者】 何荥；

【导师】 陈仲林；

【作者基本信息】 重庆大学 ， 建筑技术科学， 2008， 博士

【摘要】 在能源日趋紧张的今天,充分利用天然光,追求生态和可持续发展,节约照明用电,具有十分重要的意义。建筑物利用天然光最重要的因素是天空亮度分布,因此天空亮度分布规律就成为了天然采光的主要研究对象。本文在较为系统的分析前人有关天空亮度分布模型的基础上,利用信息法来研究天空亮度分布模型,理论上天空亮度分布模型仅与天顶亮度转换系数有关,而天顶亮度转换系数仅取决于太阳高度角的大小,与地域无关,具有普适性。也就是说在理论上仅需要一个地区的数据就可以了,这为我们研究天空亮度分布提供了一种新的方法和新的思路,其结果将会对建筑采光设计、建筑节能及充分合理的利用太阳能产生极大的影响。本文从大气辐射的基础理论出发,得到了天空亮度来源于太阳辐射在大气中的散射和吸收及影响天空亮度的主导参数是消光系数(或光学厚度)、单散射反照率以及相函数的重要结论,从而为研究天空亮度分布奠定了理论基础。参考天空是采用尽量少的天空类型来代表变化众多的真实天空,方便采光设计和采光计算,本文对各种不同气候状态下的参考天空与实际天空状况进行了比较分析,明确了参考天空与真实天空之间的关系。通过对影响天空亮度的大气光学参量进行分析研究,并在此基础上提出大气等效模型假设,提出信息法研究天空亮度的数学表达式,同时确定了该表达式中各参量的特性及在不同气候条件下的变化范围和趋势。这为天空分类及利用信息法研究天空亮度分布提供了基础与前提。本文利用多年积累的多城市、大规模光气候观测试验数据,通过对天空气象参量对天空亮度分布的影响研究,提出参考天空分类标准,即通过太阳日面状况、天空状况(云量)及地平到天顶亮度渐变状况,将天空类型划分为六类。这种分类方法具有易于观察、容易区分,能较准确的反映天空类型,并避免了现有CIE一般天空亮度标准分类模糊不清的缺陷。同时对信息法参考天空模型调整系数a、b、c、d及e对天空亮度分布的影响进行研究分析,得到不同天空类型条件下的大致取值范围,为信息法研究天空亮度模型提供了基础参数。本文利用了一种崭新的求解复杂优化问题的启发式算法——蚁群算法来研究天空亮度模型,蚁群算法具有智能搜索、全局优化、稳健性强、分布式计算、易与其它方法结合等优点。通过蚁群算法对天空亮度各影响要素的分析优化处理,得到天空亮度相对分布对表达式的各系数值。同时还开发了一套方便实用的天空亮度模型计算软件,提供直观而方便的计算手段。本文还研究了数码相机测试天空亮度的测试方法;运用经标定后的带鱼眼镜头的数码相机对天空半球拍摄测试,并采用自行开发的图像处理系统对天空亮度数据信息提取记录。通过实际测量数据与由信息法获得的理论模型的计算值相比较,其误差较小,这无疑为天然光利用,特别是建筑采光设计的高精度和标准化工作提供了无论理论上还是运用上都非常有价值的结果。论文的创新之处在于:①论文提出将天空类型分为六类的新分类标准,新分类标准解决了现有CIE标准在描述天空特征时所使用的模糊分类语言,方便了实际运用。②首次利用了一种崭新的求解复杂优化问题的启发式算法——蚁群算法来研究天空亮度模型,优化求解出天空模型各参数值,并在此基础上开发出天空模型计算系统,经验证该模型具有较高的精度。③研究利用带鱼眼镜头的数码相机和亮度计组成的图像处理系统,可方便地进行天空亮度测量,获取天空亮度测试中亮度分布数值。该系统方便易用,且有较高的精度。更多还原

【Abstract】 Make full use of natural light, pursue ecosystem and sustainable development and save lighting power is most important for the energy gradually lacks now. The most important factor of the building makes use of natural light is the sky luminance distribution, so the main study of daylight is the law of sky luminance distribution.In this paper, base on the systematic analysis of the sky luminance distribution model developed by predecessors, studying the sky luminance distribution model by information methods, the sky luminance distribution model only related to the zenith luminance conversion factor. And the zenith luminance conversion factor only depends on solar altitude angle, not concerned in geographical. It has universal application. In other words, in theory we only need a region of data. This provides a new approach and new ideas for the study of distribution of sky luminance, and it will have a grate affect on the building daylighting design, building energy conservation, adequate and reasonable use of solar energy.From the atmosphere radiation theory, this paper draw the important conclusion that sky luminance root solar radiation in atmosphere scattering and absorption and impact of sky luminance is the dominant parameters of extinction coefficient (or optical thickness), the single-scattering albedo and the phase function for study the sky luminance distribution laid a theoretical foundation.Reference sky is used to minimize the sky types to represent the fantasticality real sky, and facilitate the daylighting design and calculation. The paper analysis the reference sky and real sky in the different climate conditions, and know the relationship between the real sky and the reference sky. Through analysis the atmospheric optical parametric which impact of sky luminance, the paper bring forward the atmosphere equivalent model assumptions, getting the mathematical expression of sky luminance model by information methods, gaining the various parameters characteristics of the expression and scope of the changes and trends under different weather conditions. This provides the basis and premise for sky classification and study sky luminance distribution by information methods.The paper use the multi-city and large-scale daylihting-Climate Observing test-data accumulated over the years, through the sky meteorological parameters study on the impact of sky luminance distribution, bring forward the sky classification standard of reference skies, that is, through the Gradually change the situation of surface conditions on the sun, the sky conditions (cloud cover) and the horizon to the zenith luminance, the paper divide the sky into six types. The classification method is helpful to observation and differences, and more accurate reflection of the sky type, avoiding the dim defect of general sky luminance standard classification existing CIE. This paper analysis that the references skies model adjustment coefficient a, b, c, d and e impact the sky luminance distribution by information methods, getting value range in different sky conditions and the foundation parameters study on the sky luminance model.This paper uses a new heuristic arithmetic to solve complex optimization problem—ant colony algorithm to study the sky luminance distribution model, ant colony algorithm have many advantages ,such as intelligent search, global optimization, robustness, distributed computing and easy to integrate with other methods. Through the ant colony algorithm optimization processes infection elements of the sky luminance, we get the coefficient value of the expressions of sky luminance relative distribution. At the same time we have also developed a set of convenient and practical sky luminance model calculation software, It can provide an intuitive and easy calculation.The paper also study the testing methods of digital camera measure sky luminance, through the digital camera with fisheye lens take photo sky hemisphere, using developed by ourselves image processing system extracting records from sky luminance data information. Through compared the actual measurement data and the calculated value of theoretical model gaining by information method, its error is lesser. This is undoubtedly very valuable for daylighting utilization and especially for daylighting design of high precision and standardization in theory and application.更多还原