Article

Achieving realtime daylight factor computation for modular buildings in generative design

Authors:
To read the full-text of this research, you can request a copy directly from the author.

Abstract

In generative design, it is imperative for an architect to evaluate very quickly the performance of many buildings produced. Knowing in interactive time the daylighting potential of a generated form at an early stage of its design, with a minimum of parameters, allows to quickly choose among many variants. The daylight factor computational metamodel presented here in the case of modular buildings allows to instantly compare these solutions in order to make judicious choices in dimensioning, without performing time-consuming simulations. Another challenge was to achieve realtime computation for the daylight factor without using a GPU. We have addressed this objective via an hybrid computation both based on physical and statistical modeling, and on a physical-based computation engine specifically used for the optimization of buildings composed of multiple living units. We detail the full implementation in a generative design software leading to impressive computation times of the order of one ms.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

Article
Full-text available
Application of machine learning methods as an alternative for building simulation software has been progressive in recent years. This research is mainly focused on the assessment of machine learning algorithms in prediction of daylight and visual comfort metrics in the early design stages and providing a framework for the required analyses. A dataset was primarily derived from 2880 simulations developed from Honeybee for Grasshopper. The simulations were conducted for a side-lit shoebox model. The alternatives emerged from different physical features, including room dimensions, interior surfaces’ reflectance factor, window dimensions, room orientations, number of windows, and shading states. Five metrics were applied for daylight evaluations, including useful daylight illuminance, spatial daylight autonomy, mean daylight autonomy, annual sunlit exposure, and spatial visual discomfort. Moreover, view quality was analyzed via a grasshopper-based algorithm, developed from the LEED v4 evaluation framework. The dataset was further analyzed with an artificial neural network algorithm. The proposed predictive model had an architecture with a single hidden layer consisting of 40 neurons. The predictive model learns through a trial and error method with the aid of loss functions of mean absolute error and mean square error. The model was further analyzed with a new set of data for the validation process. The accuracy of the predictions was estimated at 97% on average. The View range metric in the quality view assessment, mean daylight autonomy and useful daylight illuminance had the best prediction accuracy among others respectively. The developed model which is presented as a framework could be used in early design stage analyses without the requirement of time-consuming simulations.
Article
Full-text available
Daylight harvesting is a well-known strategy to address building energy efficiency. However, few simplified tools can evaluate its dual impact on lighting and air conditioning energy consumption. Artificial neural networks (ANNs) have been used as metamodels to predict energy consumption with high precision, few input parameters and instant response. However, this approach still lacks the potential to estimate consumption when there is daylight harvesting, at the ambient level, where the effect of orientation can be noted. This study investigates this potential, in order to evaluate the applicability of ANNs as a tool to aid the architectonic design. The ANNs were approached as metamodels trained based on EnergyPlus thermo-energetic simulations. The network configuration focused on determining its simplest feasible form. The input parameters adopted as the main variables of the building envelope were as follows: orientation, window-to-wall ratio and visible transmission. The effects of the encoding of orientation as a network input parameter, the number of examples of each variable for network training and changing the parameters used for the training were evaluated. The networks predicted the individualized consumption according to the end use with errors below 5%, indicating their potential to be applied as a simplified tool to support the design process, considering the elementary variables of the building envelope. The discussion of results focused on guidelines and challenges to achieve this purpose when contemplating the broadening of the metamodel scope.
Article
Full-text available
Based on a new approach for the prediction of the Daylight Factor (DF), using existing empirical models, this research work presents an optimization of window size and daylight provided by the glazed apertures component for a building located in a hot and dry climate. The new approach aims to improve the DF model, considering new parameters for daylight prediction such as the orientation, sky conditions, daytime, and the geographic location of the building to fill in all the missing points that the standard DF, defined for an overcast sky, presents. The enhanced DF model is considered for the optimization of window size based on Non dominated Sorting Genetic Algorithm (NSGA II), for heating and cooling season, taking into account the impact of glazing type, space reflectance and artificial lighting installation. Results of heating and cooling demand are compared to a recommended building model for hot and dry climate with 10% Window to Wall Ratio (WWR) for single glazing. The optimal building model is then validated using a dynamic convective heat transfer simulation. As a result, a reduction of 48% in energy demand and 21.5% in CO2 emissions can be achieved. The present approach provides architects and engineers with a more accurate daylight prediction model considering the effect of several parameters simultaneously. The new proposed approach, via the improved DF model, gives an optimal solution for window design to minimize building energy demand while improving the indoor comfort parameters.
Article
Full-text available
The present study shows the application of various software performance analysis of daylighting and its energy savings potential in Energy-efficient building topologies. It can easily predict the daylight illuminance and its consequences in any energy-efficient building irrespective of any construction expense with the application of simulation software such as MatLab, Energy Plus, Velux, Relux, DIALux, in the energy analysis process. There are various computer-based tools accessible to the precise appraisal of the building's daylighting and thermal execution. This simulation software performs energy calculation for choosing effective retrofits of energy-efficient building. The study critically reviews and compares software used for energy simulation and daylighting in the building. The survey from this review found that dimming controlled daylighting committed 13% more saving of energy potential in factory building than on/off control system. The importance of daylight software in building energy saving is represented through the case study. The simulated and measured indoor illuminances with and without blinds are represented. The calculated MBE and RMSE (error) by Radiance and Radlink were satisfying each other. The modern state art of review would be helpful for the academician, scientist, and researcher working in the field daylighting.
Article
Full-text available
Steep increases in air temperatures and CO2 emissions have been associated with the global demand for energy. This is coupled with population growth and improved living standards that encourages the reliance on mechanical acclimatization. Lighting energy alone is responsible for a large portion of total energy consumption in office buildings; and the demand for artificial light is expected to grow in the next years. One of sustainable approaches to enhance energy-efficiency is to incorporate daylighting strategies, which entail the controlled use of daylight inside buildings. Daylight simulation is an active area of research that offers accurate estimations, yet requires a complex set of inputs. Even with today’s computers, simulations are computationally expensive and time-consuming, hindering to acquire accelerated preliminary approximations in acceptable timeframes, especially for the iterative design alternatives. Alternatively, predictive models that build on machine learning algorithms have granted much interest from the building design community due to their ability to handle such complex non-linear problems, acting as proxies to heavy simulations. This research presents a review on the growing directions that exploit machine learning to rapidly predict daylighting performance inside buildings, putting a particular focus on scopes of prediction, used algorithms, data sources and sizes, besides evaluation metrics. This work should improve architects’ decision-making and increase the applicability to predict daylighting. Another implication is to point towards knowledge gaps and missing opportunities in the related research domain, revealing future trends that allow for such innovative approaches to be exploited more commonly in Architectural practice.
Conference Paper
Full-text available
Building interiors daylighting is a crucial aspect for occupant comfort and energy efficiency. Standards and requirements exist to guarantee minimum levels of natural illumination in new buildings on the basis of different metrics. Some rely only on interiors static characteristics, others additionally take into account location climate. The main aim of the present study is to investigate the validity of the static Daylight Factor (DF) requirement of the Estonian daylight standard against more reliable climate-based daylight simulations using the approved method Spatial Daylight Autonomy. Results show the weakness of DF in predicting appropriate daylight availability of building interiors for the city of Tallinn. Evaluations of DF in relation to overheating regulation is also presented.
Article
Full-text available
This paper presents the latest version of our software, EcoGen2.1, to the Building Simulation community: a quite advanced prototype for generative eco-design, resulting from a French ANR project, having been the subject of a deposit with legal protection, and recently used to assist the architectural design of an emblematic building in a competition in France. The tool is adapted to the sketch step of the design, when it comes, in a given urban and climatic context, to composing the general forms of a building or a plot of buildings, depending on various criteria that ensure a sufficient level of efficiency in view of bioclimatic and environmental requirements, notably relating to solar energy, thermal consumption and luminous comfort. We have focused on the technical work involved in breaking major difficulties when trying to develop real time and interactive software devoted to generative eco-design in the early stages of an architectural project.
Conference Paper
Full-text available
Access to acceptable levels of daylight are important for people’s quality of life. Multi-unit residential buildings (MURBs) are known to perform poorly in terms of daylight compared to other residential building types. There are neither appropriate agreed upon metrics, nor effective methods for designing for daylight in MURBs. This paper presents results of a speculative design study that utilizes generative design to explore alternative geometries for the MURB tower typology. The experiment combines a genetic algorithm for spatial variation with climate-based daylight modeling (CBDM) to test the new forms against variations of the point tower MURB floorplan are commonly used. This paper identifies the poor performance of typical MURBs for daylight, and proposed new techniques for form generation. A new workflow has been developed and tested and a number of challenging issues have been identified.
Article
Full-text available
This paper presents an Artificial Neural Network (ANN) based approach able to predict the internal lighting conditions in a working environment, taking into account the daylight entering the respective space as well as the special requirements of each user. The model training procedure is based both on real illuminance and occupancy data (measurements throughout a year) and on simulations, in order to integrate all possible conditions. User preferences in respect to lighting and blinds are expressed through probability curves. Illuminance due to the external daylight is calculated and predicted throughout the whole year, depending on the weather conditions, the time of the day, the location and the office orientation. The work plane distance from the window and the usage of blinds are also considered. The proposed model is further implemented for the prediction and evaluation of energy consumption for lighting in a working space based on the user preferences.
Conference Paper
Full-text available
The climate based Daylight Autonomy (DA) metric has been gaining ground in the field of sustainable building design as a measure for the amount of daylight within spaces and associated energy savings. In this study, Artificial Neural Networks (ANNs) were used to predict DA levels in interior spaces as an alternative to computationally expensive simulations. Research was carried out in three phases of increasing complexity: First, a neural network was trained and validated for a single design space. Subsequently, the window design was altered and a neural network was trained and tested on its ability to predict DA levels according to changes in window design. Lastly, the neural network was trained to account for the effects of shading from an external obstruction. After sufficient training, the ANN, during the recall stage, was able to predict DA, on average, within 3 DA short of the simulated DA results for both the shaded and unobstructed scenario. The results obtained show the potential of neural networks as a prediction tool for estimating Daylight Autonomy.
Conference Paper
Full-text available
The aim of this paper is to examine the new method of obtaining the simulation-based results using backpropagation of errors artificial neural networks. The primary motivation to conduct the research was to determine an alternative, more efficient and less time-consuming method which would serve to achieve the results of daylight simulations. Three daylight metrics: Daylight Factor, Daylight Autonomy and Daylight Glare Probability have been used to verify the reliability of applying the latter into an early stage architectural design process. The framework was based on the computationally generated data sets build on various office models variants followed by daylight simulations. In order to predict the simulations values based on the given office parameters with artificial neural networks algorithm, a specific tool was designed as an alternative to computer simulations. The designed tool and simulations results were compared against computing time and values differences. The above findings of the research proof the reliability of the new methods as a tool during an early-stage architectural design process likewise conventional daylight simulations.
Article
Full-text available
This paper presents the findings of a research study on the development of rules-based algorithmic methods for the design of solar envelopes and optimised shapes in an urban context. They are generated using voxels according to direct solar access rules. Unlike other research projects focused on cellular automata approach, the described methods include parameters and objectives related to an eventual implementation on-site such as constructability and comfort criteria, e.g. building thickness, neighbourhood proximity, compactness, building footprint, among others. As computational and environmental tools become an integral part of the design process of architectural projects, the research opens the discussion on the voxels-based generative design and their possible impact on future urban morphologies. More specifically, subtractive and additive methods are explored and compared to generate both solar envelopes and building shapes according to quantified solar access criteria that is already applied in some countries as a requirement for light and thermal comfort.
Article
Full-text available
This paper presents three different topologies of feed forward neural network (FFNN) models for generating global, direct, and diffuse hourly solar irradiance in the city of Fez (Morocco). Results from this analysis are crucial for the conception of any solar energy system. Especially, for the concentrating ones, as direct component is seldom measured. For the three models, the main input was the daily global irradiation with other radiometric and meteorological parameters. Three years of hourly data were available for this study. For each solar component’s prediction, different combinations of inputs as well as different numbers of hidden neurons were considered. To evaluate these models, the regression coefficient ( R² ) and normalized root mean square error (nRMSE) were used. The test of these models over unseen data showed a good accuracy and proved their generalization capability (nRMSE = 13.1%, 9.5%, and 8.05% and R = 0.98, 0.98, and 0.99) for hourly global, hourly direct, and daily direct radiation, respectively. Different comparison analyses confirmed that (FFNN) models surpass other methods of estimation. As such, the proposed models showed a good ability to generate different solar components from daily global radiation which is registered in most radiometric stations.
Conference Paper
Full-text available
Current climate-based daylighting metrics have limited applicability for residential use cases and fail to highlight relevant aspects of natural light in residential spaces, such as diurnal and seasonal availability of daylight and access to direct sunlight. This paper proposes a new climate-based, annual evaluation framework that quantifies daylight autonomy and access to direct light in diurnal and seasonal bins for temperate and cold climates. Spatial maps, as well as apartment scores, can be computed. Rigorous testing at various architectural scales highlights the usefulness and sensitivity of the introduced framework.
Conference Paper
Full-text available
Architects must consider an entire year's worth of solar positions and climate data to design buildings with adequate daylight and minimal glare. However, annual simulations are time-consuming and computationally expensive, which makes them difficult to integrate into iterative design processes. In this paper, we compare the performance of several RADIANCE-based dynamic daylighting simulation methods, DAYSIM and the three-and five-phase methods and, perhaps more importantly, the potential to speed them up using parallel implementations on graphic processors. Using a model of a generic office, we achieve speedups of ten times with DAYSIM and twenty-five times with the five-phase method. Parallel implementations of three-and five-phase methods provide better scaling to multi-GPU environments and more accurate results for complex fenestration systems than parallelized DAYSIM.
Conference Paper
Full-text available
Real-time visual feedback can greatly assist the process of identifying and understanding complex cause and effect relationships. Obviously not all physical processes in the performance analysis of buildings can be simulated in real-time, but some can. By investigating these and combining them with modern games technology, it is possible to create highly visual and interactive design tools that allow all the governing parameters of a simulation, including geometry, to be manipulated by the user in real-time with detailed contextual results updating dynamically with each interactive change. This paper introduces a web-based daylighting simulation tool that does just that, using variants of the daylight coefficients and split-flux methods implemented on the GPU to calculate the real-time spatial distribution of daylight factors across a simple rectangular room. This tool's purpose is primarily educational, allowing users to gain a comprehensive understanding of the relationships between room dimensions, window placement and daylight distribution through a process of deliberate investigative play. A detailed parametric comparison with Radiance simulations resulted in a small modifier to the implemented method that produces a robust correlation which the authors argue makes this tool a valuable educational resource with potential applications in early design decision-making.
Conference Paper
Full-text available
Today's predictions of visual comfort are based on high-quality physically-based visualization renderings. Unfortunately, designers and practitioners rarely realize the full benefit of physically-based lighting simulation due to the amount of time required for these simulations. Visual comfort analysis is generally performed late in the design process as a form of validation, if at all. We propose a design workflow wherein certain quantitative visual comfort metrics can be displayed immediately to the designer as the scene changes, often before the physically-based visualization reaches a finished quality. In our prototype software, live-updating predictions of daylight glare probability, task luminance, and contrast are presented alongside a progressively rendered image of the scene so that the user may decide when to accept the values and move on with the design process. In most cases, sufficiently accurate results are available within seconds after rendering only a few frames.
Conference Paper
Full-text available
This paper examines the feasibility of implementing core algorithms from Radiance using a new type ray tracing engine optimized for highly parallel graphics hardware environments. It presents solutions to a number of implementation challenges. First, the Radiance data format is re-interpreted as a set of buffered data arrays compatible with graphics processing unit (GPU) memory. Second, the ray tracing core of the Radiance RPICT and RTRACE programs for global illumination calculations of scenes and discrete sensors is broken up into a number of small GPU programs that execute in parallel. Third, command-line user settings are declared as variables on the GPU with scopes appropriate to their functions. As a proof of concept, the paper presents a reference implementation using the OptiX™ ray tracing engine that produces images indistinguishable from Radiance up to twenty times faster for scenes with a palette of common materials.
Conference Paper
Full-text available
As people increasingly work in indoor environments, the need to provide natural lighting is becoming more widely recognized. Recent modelling standards such as LM-83 require the use of climate-based metrics based on daylight coefficients, rather than illuminance-based metrics that simulate single points in time. While calculations based on daylight coefficients are fast, computation of the daylight coefficients themselves is a slow process that must be repeated whenever the scene's geometry or materials change. Therefore, it remains impractical to obtain accurate annual daylight simulation results during early design stages when designs are fluid and quickly changing. This paper describes the development of a new, faster tool to compute daylight coefficients using graphics hardware. The tool is an adaptation of rtrace_dc, the executable used by DAYSIM to calculate daylight coefficients, written using OptiX™, a free library for GPU-based ray tracing. The effectiveness of the new tool is demonstrated using models of a typical office space with speed and results compared to rtrace_dc. The speed of the new tool is measured both on a typical workstation graphics card and on a high-end graphics server. The results show that the new tool achieves similar accuracy to the serial version but does so in one-fifth the time.
Article
Full-text available
A prediction model was developed to determine daylight illuminance for the office buildings by using artificial neural networks (ANNs). Illuminance data were collected for 3 months by applying a field measuring method. Utilizing weather data from the local weather station and building parameters from the architectural drawings, a three-layer ANN model of feed-forward type (with one output node) was constructed. Two variables for time (date, hour), 5 weather determinants (outdoor temperature, solar radiation, humidity, UV index and UV dose) and 6 building parameters (distance to windows. number of windows, orientation of rooms, floor identification, room dimensions and point identification) were considered as input variables. Illuminance was used as the output variable. In ANN modeling, the data were divided into two groups; the first 80 of these data sets were used for training and the remaining 20 for testing. Microsoft Excel Solver used simplex optimization method for the optimal weights. The model's performance was then measured by using the illuminance percentage error. As the prediction power of the model was almost 98%, predicted data had close matches with the measured data. The prediction results were successful within the sample measurements. The model was then subjected to sensitivity analysis to determine the relationship between the input and output variables. NeuroSolutions Software by NeuroDimensions Inc., was adopted for this application. Researchers and designers will benefit from this model in daylighting performance assessment of buildings by making predictions and comparisons and in the daylighting design process by determining illuminance.
Conference Paper
Full-text available
Recent developments in integrated circuit technology tend toward increased numbers of cores rather than faster clock speeds, so software must use parallelism to achieve faster run times. The ray tracing performed by Radiance is highly parallelizable in concept, with the exception of irradiance caching that serially stores and retrieves results of expensive indirect irradiation computations. This paper describes a novel method of parallel irradiance caching for global illumination on a graphics processing unit (GPU). The irradiance caching method in this paper closely follows the placement and distribution of rays produced by Radiance using the free OptiX ray tracing engine. Ambient values are stored in GPU memory which can as necessary be used to create a bounding volume hierarchy (BVH) of known irradiance records. Queries into the irradiance cache are performed by casting a short ray into the BVH. The current implementation is able to generate images similar to those created by Radiance's RPICT program up to twenty times faster.
Article
Full-text available
Nowadays, heating energy demand has become a significant estimator used during the design stage of any new building. In this paper we are proposing a model to predict the heating energy demand, based on the main factors that influence a building's heat consumption. It was found out that these factors are: the building global heat loss coefficient (G), the south equivalent surface (SES) and the difference between the indoor set point temperature and the sol-air temperature. In the second part of this paper, multiple dynamic simulations were carried out in order to determine the values of the inputs and output data of the future prediction model. Using the obtained database, a multiple regression prediction model was further used to develop the prediction model. In the last part of this paper the model results was validated with the measured data from 17 blocks of flats. Moreover, in this article it is also shown the comparison with the results calculated using the building's energy certification methodology. A detailed error analysis showed that the model presents a very good accuracy (correlation coefficient of 0.987).In conclusion, the proposed model presents the following characteristics: three inputs and one output, simplicity, large applicability, good match with the simulations and with the energy certification calculations, human behavior correction.
Article
Full-text available
The present research work concerns development of regression models to predict the monthly heating demand for single-family residential sector in temperate climates, with the aim to be used by architects or design engineers as support tools in the very first stage of their projects in finding efficiently energetic solutions. Another interest to use such simplified models is to make it possible a very quick parametric study in order to optimize the building structure versus environmental or economic criteria. All the energy prediction models were based on an extended database obtained by dynamic simulations for 16 major cities of France. The inputs for the regression models are the building shape factor, the building envelope U-value, the window to floor area ratio, the building time constant and the climate which is defined as function of the sol-air temperature and heating set-point. If the neural network (NN) methods could give precise representations in predicting energy use, with the advantage that they are capable of adjusting themselves to unexpected pattern changes in the incoming data, the multiple regression analysis was also found to be an efficient method, nevertheless with the requirement that an extended database should be used for the regression. The validation is probably the most important level when trying to find prediction models, so 270 different scenarios are analysed in this research work for different inputs of the models. It has been established that the energy equations obtained can do predictions quite well, a maximum deviation between the predicted and the simulated is noticed to be 5.1% for Nice climate, with an average error of 2%. In this paper, we also show that is possible to predict the building heating demand even for more complex scenarios, when the construction is adjacent to non-heated spaces, basements or roof attics.
Article
Full-text available
The aim of this research work is to propose sky radiance and luminance distribution models for all sky conditions from clear sky to overcast sky. The classification method of sky conditions is examined. Clear sky index is defined based on the global irradiance. Cloudless index is defined based on the global irradiance and the diffuse irradiance. The sky conditions are classified according to both indices. The data of the average sky radiance (luminance) distributions are obtained based on the classified sky conditions. The average relative sky radiance distributions are compared with the average relative sky luminance distributions. It has been confirmed that both radiance/luminance distributions can be shown by the same equation.An equation that shows the relative sky radiance (luminance) distribution for all sky conditions and equations that show the zenith radiance and the zenith luminance are obtained. The absolute value of the sky radiance (luminance) distribution is shown by multiplying the relative sky radiance (luminance) distribution and the zenith radiance (luminance). The equations to show the absolute values of the sky radiance distribution called the All Sky Model-R and the sky luminance distribution called the All Sky Model-L are proposed.Both models are compared with the previous models based on the measurement distributions concerning the region of sky vault and the sky conditions and good results are obtained. In addition, the vertical irradiance and illuminance calculated by the proposed models and the previous models are compared with the measurements, and the proposed models are confirmed their accuracy.
Conference Paper
Full-text available
The OptiX™ engine is a programmable ray tracing system designed for NVIDIA® GPUs and other highly parallel architectures. OptiX builds on the key observation that most ray tracing algorithms can be implemented using a small set of programmable operations. Consequently, the core of OptiX is a domain-specific just-in-time compiler that generates custom ray tracing kernels by combining user-supplied programs for ray generation, material shading, object intersection, and scene traversal. This enables the implementation of a highly diverse set of ray tracing-based algorithms and appli- cations, including interactive rendering, offline rendering, collision detection systems, artificial intelligence queries, and scientific sim- ulations such as sound propagation. OptiX achieves high perfor- mance through a compact object model and application of several ray tracing-specific compiler optimizations. For ease of use it ex- poses a single-ray programming model with full support for recur- sion and a dynamic dispatch mechanism similar to virtual function calls.
Article
Full-text available
We present a new method for solving stochastic differential equations based on Galerkin projections and extensions of Wiener's polynomial chaos. Specifically, we represent the stochastic processes with an optimum trial basis from the Askey family of orthogonal polynomials that reduces the dimensionality of the system and leads to exponential convergence of the error. Several continuous and discrete processes are treated, and numerical examples show substantial speed-up compared to Monte Carlo simulations for low dimensional stochastic inputs.
Article
Daylight analysis is essential in building design to ensure indoor environment quality, including health and thermal comfort vis-à-vis energy. It is a repeating and time-consuming process of design options. Several studies conducted machine learning models to accurately predict daylight performance in particular design situations. Therefore, developing an AI-based real-time daylight analysis platform becomes more promising. However, buildings can be designed with arbitrary shapes, creating a real challenge for the AI to recognize any building layout. From that perspective, the idea of finding the design variables that characterize all the building layouts becomes the key solution. To unlock this challenge, we promote a novel method of creating design variables and building a machine learning model that can efficiently forecast daylight performance with different building layouts. The daylight metric was Useful Daylight Illuminance with four ranges, and the case studies were assumed medium-sized buildings located in Ho Chi Minh City, Vietnam. All the data for training and predicting were created by the simulation DIVA tool. Obtained results showed the excellent performance of the proposed approach, which brings more promising in developing a data-driven machine learning platform for real-time daylight validation. Moreover, the present framework can adapt to any specific machine learning model or daylight simulation tool and daylight metrics.
Article
In order to obtain information about building daylighting performance earlier in the design stage, recent studies have begun to develop daylight prediction models using machine learning methods. Most studies have adopted design variation parameters as the input parameters for model training, but this method greatly limits the scope of application of the daylight model. By extending application to different design possibilities, this research proposes a novel daylight model, which includes a pre-processing procedure to convert the design model into “Intermediary features” as input parameters representing daylight penetration performance. By changing geometry of one kind of parametric façade and performing daylighting simulation, we were able to generate the data for training, including Intermediary features and sDA/ASE values. The daylight model was then trained using an artificial neural network. Finally, the proposed daylight model was tested by predicting the daylight performance of other kinds of façades. The results indicated that the value of DA and ASE hours per grid were well predicted and that the daylight distribution was reproduced even with different kinds of façades. The deviations of sDA and ASE from simulation results ranged from 1.7 to 6.1% and 0.3–2.1 h respectively. The reproducibility, the predictive capability, and most importantly the extension of model applicability were all demonstrated for the proposed model. Furthermore, comparing to the daylighting simulation, the method using the proposed daylight model is estimated to save 9/10 daylighting evaluation time. This is critical for implementing the evaluation in the early design stage.
Conference Paper
With the innovation of technology, both our lit environment and the way people perform indoor tasks have changed. Good visual performance became relatively easy to achieve, and as a result the emphasis of lighting design has moved away from the lighting of working planes. Whilst task illuminance is still in use there is now much more emphasis on the appearance of the room and the people in them. In fact, the term “working plane” has been nominally removed in the European electrical lighting standard. Therefore, it is necessary to question the use of working planes in daylight designs. For many years daylight factor has been the dominant metric used to describe the amount of daylight in a room. However, it only considers light falling onto the working plane and thus it may not be the best metric to describe daylight adequacy in modern buildings. A few alternative metrics have been proposed such as the metrics by Climate-based daylight modeling (DA, UDI) which recently have raised a lot interests. In addition, a number of new lighting parameters (MRSE, cubic illuminance and cylindrical illuminance) have also been proposed but only applied to electrical lighting. This research studied a new metric for daylight, derived from MRSE, to find out if the new metric was better at predicting user perceptions of daylight adequacy than the existing working plane based metrics.
Article
Fueled by big data, powerful and affordable computing resources, and advanced algorithms, machine learning has been explored and applied to buildings research for the past decades and has demonstrated its potential to enhance building performance. This study systematically surveyed how machine learning has been applied at different stages of building life cycle. By conducting a literature search on the Web of Knowledge platform, we found 9579 papers in this field and selected 153 papers for an in-depth review. The number of published papers is increasing year by year, with a focus on building design, operation, and control. However, no study was found using machine learning in building commissioning. There are successful pilot studies on fault detection and diagnosis of HVAC equipment and systems, load prediction, energy baseline estimate, load shape clustering, occupancy prediction, and learning occupant behaviors and energy use patterns. None of the existing studies were adopted broadly by the building industry, due to common challenges including (1) lack of large scale labeled data to train and validate the model, (2) lack of model transferability, which limits a model trained with one data-rich building to be used in another building with limited data, (3) lack of strong justification of costs and benefits of deploying machine learning, and (4) the performance might not be reliable and robust for the stated goals, as the method might work for some buildings but could not be generalized to others. Findings from the study can inform future machine learning research to improve occupant comfort, energy efficiency, demand flexibility, and resilience of buildings, as well as to inspire young researchers in the field to explore multidisciplinary approaches that integrate building science, computing science, data science, and social science.
Chapter
Since shape grammars were first described about forty five years ago, several types of grammars have emerged. The goal of this paper is to provide a framework that gathers together the main existing types of grammars. The categorization is preceded by a glossary of 19 terms related to shape grammars. Then, 44 types are placed into 13 chronologically-ordered categories. Each type is characterized with its name, a short description, the reference to the original paper, three examples of existing grammars of the type, and simple illustrative grammars. The types are organized using a classification guide in the form of a checklist, which is filled according to an existing shape grammar as an example.
Article
Prediction of the daylight level for any point within an indoor space is important in daylighting analysis. Recently, the International Commission on Illumination (CIE) has adopted a range of 15 standard skies covering the whole probable spectrum of usual skies found in the world. In densely-built urban cities, the exterior obstructions may substantially reduce the amount of daylight entering to the building interior. Through exhaustive investigations, we proposed various daylight computation tools for buildings under different obstructed overcast and non-overcast skies. This paper reviews the daylight factor (DF) based metrics and gives an overview of the latest dynamic daylight metrics, and generalizes the calculation procedures for estimating the internal daylight under the unobstructed and obstructed 15 CIE Standard Skies. The merits and limitations of the calculating procedures on DF-based metrics are discussed. The findings could be globally adopted and useful to students and building professions conducting daylighting studies and designs.
Article
This paper represents an illumination modeling method for lighting control which can model the illumination distribution inside office buildings. The algorithm uses data from the illumination sensors to train Radial Basis Function Neural Networks (RBFNN) which can be used to calculate 1) the illuminance contribution from each luminaire to different positions in the office 2) the natural illuminance distribution inside the office. This method can be used to provide detailed illumination contribution from both artificial and natural light sources for lighting control algorithms by using small amount of sensors. Simulations with DIALux are made to prove the feasibility and accuracy of the modeling method.
Article
Recursive ray tracing is a simple yet powerful and general approach for accurately computing global light transport and rendering high quality images. While recent algorithmic improvements and optimized parallel software implementations have increased ray tracing performance to realtime levels, no compact and programmable hardware solution has been available yet.This paper describes the architecture and a prototype implementation of a single chip, fully programmable Ray Processing Unit (RPU). It combines the flexibility of general purpose CPUs with the efficiency of current GPUs for data parallel computations. This design allows for realtime ray tracing of dynamic scenes with programmable material, geometry, and illumination shaders.Although, running at only 66 MHz the prototype FPGA implementation already renders images at up to 20 frames per second, which in many cases beats the performance of highly optimized software running on multi-GHz desktop CPUs. The performance and efficiency of the proposed architecture is analyzed using a variety of benchmark scenes.
Article
The envelope shape, the most salient design characteristic in a building, has significant influence on its energy performance. However, in the early design stages, when the envelope shape is defined, energy performance information is normally nonexistent, due to modeling for energy simulation being a time-consuming task, frequently overlooked at this phase. This paper presents a methodology to assist design decisions regarding the building envelope shape considering its implications on energy performance. Basically, this methodology involves a flexible design system, to generate alternative envelope shape designs, with integrated energy simulation, to calculate the energy demand of each design. Shape grammars are particularly suitable to encode architectural design systems, given their ability to encode compositional design principles. Their downside is the complexity in developing computer implementations. This methodology converts a grammar into a parametric design system and is illustrated with an application to the grammar for Frank Lloyd Wright's prairie houses.
Article
Solar irradiance and outdoor illuminance, particularly on vertical surfaces are crucial to energy-efficient building designs and daylighting schemes. In Hong Kong, only hourly horizontal global solar radiation data have been systematically recorded for a long period but no measurements of daylight illuminance exist. In 2003, the International Commission on Illumination (CIE) adopted a range of 15 standard skies covering the whole probable spectrum of skies in the world. Standard skies of the same category would have the identical well-defined sky radiance and luminance distributions. Once the skies are identified, the basic solar irradiance and daylight illuminance at the surfaces of interest can be obtained, involving simple mathematical expressions. This study presents a numerical approach to predict the vertical solar irradiance and daylight illuminance based on the CIE standard skies. Climatic parameters recorded between January 2004 and December 2005 are used in the analysis. The performance of the calculation method is evaluated against data measured in the same period. The annual RMSEs were found ranging from 17.7% to 20.8% for daylight illuminance prediction and 17.9%–19.8% for solar irradiance estimation. The findings provide an alternative to compute solar irradiance and daylight illuminance on vertical surfaces facing various orientations.
Article
This paper describes the application of a new paradigm, called useful daylight illuminance (UDI), to assess daylight in buildings. The UDI paradigm is designed to aid the interpretation of climate-based analyses of daylight illuminance levels that are founded on hourly meteorological data for a period of a full year. Unlike the conventional daylight factor approach, a climate-based analysis employs realistic, time-varying sky and sun conditions and predicts hourly levels of absolute daylight illuminance. The conventional approach produces a single number – the daylight factor as a percentage – for each evaluation point in the space. In contrast, a climate-based analysis results in an illuminance prediction for every daylight hour of the year for each point considered. The UDI paradigm offers a way to reduce the voluminous time-series data to a form that is of comparative interpretative simplicity to the daylight factor method, but which nevertheless preserves a great deal of the significant information content of the illuminance time-series. The UDI paradigm informs not only on useful levels of daylight illuminance, but also on the propensity for excessive levels of daylight that are associated with occupant discomfort and unwanted solar gain. In a conventional analysis of daylight provision and solar penetration, the two phenomena are assessed independently using methods that are idealised (daylight factor) and qualitative (shadow patterns). The UDI paradigm offers a simple methodology whereby daylight provision and levels of solar exposure are quantified using a single evaluative schema. Thus, it is also well-suited for teaching purposes. Application of the UDI paradigm is demonstrated using an analysis of design variants for a deep-plan building with a light-well. Comparison is made with the conventional daylight factor approach, the LEED daylight credit and measures of daylight autonomy.
Article
In urban canyons where, apartment buildings can be beneficial in terms of their close proximity to offices, shops etc, causing less traffic congestion and pollution, saving fuel costs and bringing people in close proximity to city centres. At the same time trying to accommo- date people in city centres, leads to the development of congested and confined narrow deep apartments which have to be mechanically ventilated and artificially lit. To reduce costs, in 1962 designers began building nar- row structures with light curtain walls and thin frames to increase daylight and natural ventilation. These fac- tors bought about overheating and noise pollution in buildings. More focus then went on to the design of windows and their fixtures to prevent heat loss, over - heating and internal acoustical treatment of ceilings and floors to prevent acoustical reverberations 1. The above views that windows were net energy debits disregard- ed the net solar heat gains in winters and reduction in electric lighting loads through out the year. Experience now shows that well day-lit buildings have lower run- ning costs and are more energy efficient. In restricted sites, designers tend to design spaces with maximum window areas' specifically using full height windows to compensate for deep spaces in maintaining daylight. This study lead to investigating the validity of the BRE Average daylight factor formula for how windows with increased areas below the working plane contribute to the ADF in 'Site planning,etc'4.
Article
Daylighting supports human activities with energy savings and provides positive effects to human health. In a dense urban environment, ensuring adequate daylight incident on windows is of the utmost importance in the design of building form and layout. Vertical daylight factor (VDF) (defined and explained in the text) has been used as a measure of the daylight availability on window façades. However, there is a lack of comprehensive method for the calculation of VDF. This paper presents a method of VDF calculation based on first principles. The method is also demonstrated for four building layouts with a CIE overcast sky. Results obtained by the presented method are compared with a validated lighting simulation program RADIANCE. A simplified method is also presented at the end of this paper for easy and fast estimation of VDF.
Article
Computational design techniques are changing the role of analysis tools in collaborations between architects and engineers. Digital feedback loops of synthesis, analysis and evaluation establish a ‘process of becoming’ in which structural solutions evolve and adapt to specific requirements. Highly differentiated constructions are possible when digital techniques are fully integrated in design and production. Klaus Bollinger, Manfred Grohmann and Oliver Tessmann discuss these novel paradigms in relation to recent projects from engineering office Bollinger + Grohmann. Copyright © 2010 John Wiley & Sons, Ltd.
Article
In non-residential buildings, comfort and energy demand for heating, cooling and lighting are significantly influenced by the façade. Up to now, only non-weighted luminance-based methods for calculating and evaluating annual daylight glare exist (Lee et al., 2005; Mardaljevic and Lomas., 1998). Within this paper, different methods based on the daylight glare probability DGP (Wienold and Christoffersen, 2006) for a dynamic calculation of glare are discussed and evaluated: 1. Timestep by timestep calculation – RADIANCE reference method. 2. Simplified daylight glare probability DGPs – DGP only based on vertical eye illuminance. Results of this method are similar to average luminance based evaluations. 3. Enhanced simplified DGP calculation -DGP based on vertical eye illuminance and simplified images. The enhanced simplified DGP method is validated against two hour-by-hour full year calculations, using a fabric and a Venetian blinds shading system. For the yearly evaluation of dynamic glare results, a histogram analysis and a glare rating classification is proposed. Download here: http://www.ibpsa.org/proceedings/BS2009/BS09_0944_951.pdf
Article
An artificial neural network (ANN) model for estimating sky luminance was developed. A 3-year period (2007-2009) of sky luminance data obtained from measurements at Nakhon Pathom (13.82°N, 100.04°E) and a 1-year period (2008) of the same type of data at Songkhla (7.20°N, 100.60°E), Thailand were used in this study. The ANN model was trained using a back propagation algorithm, based on 2 years data (2007-2008) at Nakhon Pathom for clear, partly cloudy and overcast skies. The trained ANN model was used to predict sky luminance at Nakhon Pathom for the year 2009 for the case of clear, partly cloudy and overcast skies. The results were compared with those of the CIE model. It was found that the ANN model performed better than CIE models for most cases. The ANN model trained with Nakhon Pathom data were also used to predict sky luminance at Songkhla and satisfactory results were obtained.
Article
Daylight illuminance, particularly on vertical surfaces, plays a major role in determining and evaluating the daylighting performance of a building. In many parts of the world, however, the basic daylight illuminance data for various vertical planes are not always readily available. The usual method to obtain diffuse illuminance on tilted planes would be based on inclined surfaces models using data from the horizontal measurements. Alternatively, the diffuse illuminance on a sloping plane can be computed by integrating the luminance distribution of the sky ‘seen’ by the plane. This paper presents an approach to estimate the vertical outdoor illuminance from sky luminance data and solar geometry. Sky luminance data recorded from January 1999 to December 2001 in Hong Kong and generated by two well-known sky luminance models (Kittler and Perez) were used to compute the outdoor illuminance for the four principal vertical planes (N, E, S and W). The performance of this approach was evaluated against data measured in the same period. Statistical analysis indicated that using sky luminance distributions to predict outdoor illuminance can give reasonably good agreement with measured data for all vertical surfaces. The findings provide an accurate alternative to determine the amount of daylight on vertical as well as other inclined surfaces when sky luminance data are available.
Article
This article reports the development and evaluation of a new model for describing, from routine irradiance measurements, the mean instantaneous sky luminance angular distribution patterns for all sky conditions from overcast to clear, through partly cloudy, skies.
Article
A class of incomplete three level factorial designs useful for estimating the coefficients in a second degree graduating polynomial are described. The designs either meet, or approximately meet, the criterion of rotatability and for the most part can be orthogonally blocked. A fully worked example is included.
A three-dimensional modular building system
  • M Safdie