Glare Prediction Research Articles

Fusing Transformer and diffusion for high-resolution prediction of daylight illuminance and glare based on sparse ceiling-mounted input

Accurate prediction of workplane daylight illuminance and eye-height glare is crucial for lighting control. Existing studies used machine learning to predict illuminance at predetermined locations based on indoor sensors, but they may encounter challenges in scenarios 1) with flexible seating arrangements, 2) with dynamic shading devices, and 3) requiring the prediction of glare. To address these challenges, we proposed a novel method fusing Transformer and Diffusion models, with the input being data collected from sparse ceiling-mounted illuminance sensors, and the outputs being high-resolution workplane illuminance and glare. The model works well for rooms without and with dynamic roller shades. For the former, the mean absolute errors for illuminance below 3000 lx and Daylight Glare Index (DGI) are only 20.77 lx and 0.20, and the error rates in detecting illuminance <500 lx and DGI>22 are only 0.85 % and 5.55 %. For the more complicated latter case, the aforementioned four numbers are 34.78 lx, 0.59, 2.47 % and 23.13 %. The model significantly outperforms the linear and the ANN models, particularly in glare prediction. The influence of sensor number and placement strategy on model performance was also revealed. The model can potentially enhance lighting control, especially in cases with dynamic shading, with flexible seating arrangements, and where glare is of interest.

Comparison of Simulation Methods for Glare Risk Assessment with Roller Shades

Daylight discomfort glare evaluation is important when selecting shading properties. New standards recommend allowable glare frequency limits but do not specify the modeling accuracy required for annual glare risk assessment. Fast simulation tools allow users to perform hourly glare evaluations within minutes. However, reliable evaluation of glare through roller shades requires accurate modeling of their specular and diffuse transmission characteristics, affected by color, materials, and weaving technology. This study presents a systematic comparison between commonly used glare simulation methods against the “ground truth” Radiance ray-tracing tool rpict in terms of hourly daylight glare probability (DGP), hourly vertical illuminance (Ev), and annual visual discomfort frequency. The results are presented for two shade fabrics using light transmission models with and without a peak extraction algorithm (Radiance–aBSDF and Radiance–BSDF) for the specular component. The impact of sky/sun discretization on glare prediction is also discussed. The results show that the Radiance 5–Phase Method (5PM) is superior when modeling direct sunlight and DGP through shades, while other investigated methods (3–Phase Method, imageless DGP, ClimateStudio Annual Glare) are not as robust for that purpose. Users are encouraged to understand the underlying assumptions in the imageless methods to avoid errors when simulating glare, especially due to the contrast effects.

Glare prediction and mechanism of adaptation following implantation of hydrophilic and hydrophobic intraocular lenses.

Glare is a known side effect of intraocular lens (IOL) implantation, affected principally by IOL material and optics, although it is reported subjectively to decrease in impact with time. However, little objective data have been published on changes over time, how these relate to subjective reports, and whether those who will report greater glare symptoms can be predicted prior to IOL implantation. A total of 32 patients (aged 72.4 ± 8.0 years) with healthy eyes were implanted bilaterally with hydrophilic 600s (Rayner, Worthing, UK) or hydrophobic Acrysof (Alcon, Texas, USA) acrylic IOLs (n = 16 each, randomly assigned). Each patient reported their dysphotopsia symptoms subjectively using the validated forced choice photographic questionnaire for photic phenomena, and halo size resulting from a bright light in a dark environment was quantified objectively in eight orientations using the Aston Halometer. Assessment was performed binocularly pre-operatively and at 1, 2, 3, and 4 weeks after IOL implantation. The study was carried out at the National Health Service Ophthalmology Department, Queen Elizabeth Hospital, Birmingham, UK. Visual acuity (average 0.37 ± 0.26 logMAR) did not correlate with subjective glare (r = 0.184, p = 0.494) or objective glare (r = 0.294, p = 0.270) pre-surgery. Objective halo size (F = 112.781, p < 0.001) decreased with cataract removal and IOL implantation and continued to decreased over the month after surgery. Subjective dysphotopsia complaints (p < 0.001) were also greater pre-surgery, but did not change thereafter (p = 0.228). In neither case was there a difference with IOL material (p > 0.05). It was not possible to predict post-surgery dysphotopsia from symptoms or a ratio of symptoms to halo size pre-surgery (p > 0.05). Subjective dysphotopsia and objective halos caused by cataracts are greatly reduced by implantation of IOL after cataract removal causing few perceivable symptoms. However, objective measures are able to quantify a further reduction in light scatter over the first month post-IOL implantation, suggesting that any subjective effects over this period are due to the healing process and not due to neuroadaptation.

Investigation of influential variations among variables in daylighting glare metrics using machine learning and SHAP

Previous studies have shown the inconsistent performance of various daylighting glare prediction metrics in office daylit-dominated environments. This lack of consensus may stem from a limited understanding of how their component variables contribute to the prediction power reflected on these discomfort glare models. Previous studies predominantly focused on individual variables' linear, independent impact on glare discomfort prediction, neglecting to explore their non-linear and interaction contributions. This study addressed these limitations by evaluating regularly used glare variables using machine learning-based feature selection and SHAP interpretation methods to parse the collected dataset in a subjective survey. By applying different machine learning algorithms and SHAP explainers, we have identified their contribution importance to the prediction of different scales of glare discomfort rating when incorporated into one model and have observed the difference of interaction effects between these variables on predicting glare discomfort ratings. Our findings indicate that vertical eye illuminance, maximal luminance, and window mean luminance exhibit strong prediction power and significant interaction effects on glare discomfort sensations, indicating that they could be the dominant contributors to glare predictive metrics. Contrast effect variables within the field of view had less predictive influence on glare scales when combined with other variables. This innovative approach enables a comprehensive assessment of different variables' influential sequence and interaction consequences when incorporated into the glare prediction model. It facilitates the development of the new glare discomfort prediction model.

An examination of reflected glare prediction method based on luminance distribution

Explicit evaluation methods and suppression techniques for reflected glare have not been clarified in museum lighting. The purpose of this study is to elucidate the occurrence mechanism of reflected glare and to make use of its knowledge for future exhibitions in museums. In this paper, for the first step, we focused on the several factors, which were assumed to have an influence on the intensity of the reflected glare, such as the balance between ambient light levels for an exhibition room and target light levels for paintings, the Munsell value of paintings, and the amount of varnish coating. We conducted a subjective experiment in which the reflected glare on the replica paintings was evaluated by subjects aged in their 20s in an experimental room that was assumed to be an actual exhibit space. Based on the results of this experiment, we confirmed that the degree of reflected glare in museums could be roughly estimated from luminance distribution, reflectance, and glossiness of the paintings.

MACHINE LEARNING MODEL FOR GLARE PREDICTION IN OFFICES WITH SIMPLE ARCHITECTURAL FEATURES

ABSTRACT Daylight glare index (DGI), daylight glare probability (DGP) and glare-sensation (GS) predictive models are the widely used glare indices for the assessment of occupant visual comfort in daylit spaces. This paper presents the development and implementation of Machine Learning models to predict these glare indices. The training and validation data sets were collected from sensors incorporated in the test room with motorized Venetian Blinds and dimmable LED luminaires. Predictor and response data were obtained from conventional sensors, digital cameras, and the EVALGLARE Software. The regression models predict DGI and DGP, whereas the classification model predicts GS. In addition to standard statistical error evaluation metrics, the hypothesis test assesses the performance of regression/classification models. The results reveal that Ensemble Tree (ET) models are highly accurate at predicting glare indices. The proposed technique attempts to simplify the existing traditional Glare Index(GI) estimation method. The combination of real-time daylight glare prediction and suitable window shading control increases occupant visual comfort. A high dynamic image-based system is employed to verify the measurements made using traditional sensors.

Discomfort glare from daylight in classrooms

Provision of daylight without the risk of discomfort glare is one of the aspects that determine the quality of the classroom environment. Although discomfort glare from daylight is under investigation for a long time, a knowledge gap concerning the applicability of the existing metrics to the spatial conditions of the classroom is identified in this work. An investigation on the applicability of existing metrics to the prediction of discomfort glare in classrooms has been carried out based on two experimental studies. The analysis shows that the existing metrics have poor predictive ability particularly in the sitting positions away from the window light source. A study is then carried out to investigate how can more appropriate predictive models of discomfort glare be developed, based on three different methods. A modification of the Daylight Glare Probability (DGP) equation produced a significantly better discomfort glare model than any of the metrics that have been studied in this work. Following this finding, a new metric, DGPlog(Ev)new, is proposed. The produced metric suggests that discomfort glare in the classroom is better identified based on a range of equations for different sitting positions or that new variables that account for sitting position need to be included in a predictive model of discomfort glare for these spaces. A set of architectural design guidelines towards a discomfort glare free classroom is then proposed, based on the newly produced model and on the collected data.

A Critical Comparison of Annual Glare Simulation Methods

Recently, there have been multiple proposals for faster methods to calculate glare metrics, daylight glare probability (DGP) in particular. This is driven simultaneously by the lengthy times required to simulate DGP with a conventional image-based approach and accumulating evidence from subjective glare evaluation experiments showing that accounting for both the saturation and contrast in a view improves the accuracy of a glare prediction. While some of these methods have been presented with their own validations, comparisons of accuracy across methods are limited by the differences in tested scenarios and resulting distributions of daylight conditions. This study compares six point-based methods and two zonal estimations for quickly calculating hourly DGP values from three viewpoints with different relationships to the window across a range of scenarios. These include scenarios with direct and semi-specular transmission and others with specular and semi-specular reflection. We find that while some of these fast methods closely align with results from a reference simulation, others introduce large and consequential errors.

Comparative Investigation of Daylight Glare Probability (DGP) Comfort Classes in Clear Sky Condition

Glare is considered one of the most important variables to reach visual comfort and visual quality. It represents one of the fundamental barriers for an effective use of daylighting in buildings. One of the best performing and robust glare prediction models, relative to other available metrics, is a Daylight Glare Probability (DGP). Based on a validated and precise methodology (RADIANCE) the aim of this work is to compare the DGP model (original cut-off values) with new cut-off values that differ according to the time of day (morning, noon and afternoon). Both cut-off values were compared at more than 300 simulated conditions of daylighting in an interior space. This work offers the originality of studying recently proposed cut-off values in climate luminous with predominant clear sky conditions. Currently, the application of these new cutoff values is reduced to the field of science or simulation professionals. The results showed important differences (64.86%) between the categories proposed by both cut-off values. Nevertheless, these differences do not have a significant impact in glare prediction (&lt; 2.7%), in terms of glare absence (DGP &lt;0.38) and presence (DGP &gt;0.38). This analysis made it possible: (i) to regionally apply the main current corpus criteria regarding glare issues as well as emergent proposals and (ii) to present new experimental data aimed at helping the field and, together with other works, improving the tools used by professionals on a daily basis.

Comparing performance of discomfort glare metrics in high and low adaptation levels

Current discomfort glare prediction metrics usually account for at least one of the two categories of effects that induce discomfort glare – the saturation and contrast effects. Saturation-driven metrics (overall illuminance on the eye) are suited for brightly lit scenes in general. On the contrary, contrast-driven metrics (luminance ratio in the field of view) usually perform better in high contrast conditions such as with small-sized bright glare sources. Only a few existing metrics consider both effects, such as the Daylight Glare Probability (DGP). However, even these “hybrid” metrics may underperform in conditions other than those considered when they were developed, such as in dim scenes with high contrast glare. This paper investigates the ability of current glare indicators to predict perceived discomfort glare in user-evaluated scenes depending on two different adaptation levels. Towards this end, we used a composite dataset of six laboratory studies performed previously and separately in various parts of the world. According to Receiver Operator Characteristics (ROC) findings and complementary statistical research, the hybrid metrics DGP and Eccologit perform best in both investigated ranges (dimmer and brighter scenes). For the single-effect metrics, the contrast-driven metrics appear to perform better than saturation-driven metrics in lower adaptation levels (dimmer scenes), while the reverse is seen in higher adaptation levels (brighter scenes). As a result, metrics that only consider one effect should be used with caution. Although hybrid metrics continue to perform well in the investigated scenes, further research is needed to extend their applicability to a larger variety of lighting conditions that may be observed in work environments.

The Usefulness of Near-Field Goniophotometry Data to Assess Illuminances and Discomfort Glare in Indoor Lighting

ABSTRACT In lighting simulations, a luminaire is usually modeled using a luminous intensity distribution emitted by a single-point source. This so-called far-field approach is valid as long as the dimensions of the luminaire are much smaller than the distance to the calculation surface. This assumption is generally not met in standard indoor lighting and may lead to significant errors in the prediction of illuminances and discomfort glare. This work describes a practical near-field approach based on splitting the luminaire into N point sources. The errors in relation to the near field (infinite value of N) are evaluated as a function of N. This approach generates a more accurate assessment of the Unified Glare Rating in the case of luminaires with non-uniform emission.

Glare prediction of solar PV system in airport environment: A scenario analysis of material, tilt and orientation

The application of solar PV is suitable in airport areas for energy generation due to the availability of vast open spaces. At the same time, there is possibility of glare from solar PV array which may affect the visibility of pilots and air traffic controllers. This paper attempts to perform glare assessment for conceptual solar PV system located in an airport of India using ForgeSolar software under four different scenarios. These four scenarios are solar PV modules with smooth glass, smooth glass with anti-reflective coating (ARC), light textured glass with ARC and different tilt & orientation angles respectively. The location of the proposed site, details of solar PV array and characteristics of observers are provided. The occurrence and duration of glare for each scenario is obtained from the software. The impact of glare is assessed based on solar glare standard set by Federal Aviation Administration (FAA). Among the studied scenarios, safe value of glare is obtained in Scenario 4 only at specific values of tilt and orientation angle. A tilt angle of 15 degrees and orientation angle of 135 degrees is the suitable value with 99.1 % of theoretical energy output. It is observed that the yellow glare impact on ATC reduced in the descending order of Scenario 1 (2122 min), Scenario 2 (296 min) and Scenario 3 (0 min). It is concluded that proper selection of tilt and orientation angle of solar PV array aids in achieving zero minutes of glare for airport conditions.

Determination of the Simplified Daylight Glare Probability (DGPs) Criteria for Daylit Office Spaces in Thailand

The increasing trend of employing glazed façades to utilize daylight in the buildings has made it necessary to develop measures to avoid excessive sunlight penetration in such daylit spaces. In Thailand, only a few studies have focused on daylight glare, and therefore, applicable criteria are required to fulfill the local preference. This study aimed to determine daylight glare thresholds on the basis of the occupants’ responses. A post-occupancy evaluation with a simplified daylight glare probability (DGPs) model was performed in eight open-plan office spaces located in Bangkok, Thailand. The occupants participated in a survey including a subjective questionnaire; the results showed that the DGPs model performed effectively for glare prediction, with a preference for a lower level than that found in the current references. Statistical analysis helped mark the threshold values for each glare sensation level: imperceptible–perceptible = 0.22; perceptible–disturbing = 0.24; and disturbing–intolerable = 0.26. The findings of this study can be considered as initial evidence for improving the understanding of local occupants’ perspectives and illumination standards, which currently encourage daylight utilization without any specific glare control strategies.

Technical assessment of captive solar power plant: A case study of Senai airport, Malaysia

Solar PV system in the airport environment is a relatively new application. Unlike land-based solar systems, the site selection for the airport-based PV power plant is a complicated process and lacks proper methodology. The objective of this work was to develop a general sitting procedure for an airport-based solar PV system and identify ideal sites for solar farms in Senai International airport, Malaysia. Feasible sites were selected with due consideration to airport and aviation compatibility constraints. Next, suitability of such selected sites is assessed based on environmental impact and proximity to electrical infrastructure. Using glare prediction software, the adherence to FAA’s solar interim policy is assessed. Eleven (11) sites which lie within the airport are chosen for the study. The duration of glare from sites 2, 3, 4, 6 were 1125, 4724, 3805, 1125 min receptively. As a result, design parameters are changed for these sites. The results of the study showed that the solar PV potential and theoretical energy generation from the selected sites of the airport were 12.50 MW and 16,745 MWh respectively. The knowledge on the suitability of sites and prior glare assessment increases the level of confidence to airport stakeholders and project developers.

Analysis of solar PV glare in airport environment: Potential solutions

Abstract Solar PV plants are being installed in many airports around the globe. Reflection from the solar PV arrays is a big concern for airport stakeholders. This paper aims to assess the glare occurrence and its impact from the proposed solar PV plant installed in an airport. The prediction of glare is carried out with the help of computational software for a randomly selected area within the boundary of Kuantan airport, Malaysia. The selected zone is not suitable for solar installations as per FAA's glare policy. In the selected area (Apv), the duration of glare on ATCT from solar modules installed is 6778 min (green and yellow glare). Also, the flight path is free from any kind of glare occurrence. Glare occurs between March to mid-October, mostly from 7.00 a.m. to 8.00 a.m. Green and Yellow glare last up to 10 min and 30 min respectively in a year. In this regard, remedial steps for mitigating possible glare are discussed. Glare prediction helps in the early adoption of suitable remedial measures against glare hazard.

Cross-validation and robustness of daylight glare metrics

This study evaluates the performance and robustness of 22 established and newly proposed glare prediction metrics. Experimental datasets of daylight-dominated workplaces in office-like test rooms were collected from studies by seven research groups in six different locations (Argentina, Denmark, Germany, Israel, Japan and the United States). The variability in experimental setups, locations and research teams allowed reliable evaluation of the performance and robustness of glare metrics for daylight-dominated workplaces. Independent statistical methods were applied to individual datasets and also to one combined dataset to evaluate the performance and robustness of the 22 glare metrics. As performance and robustness are not established in literature, we defined performance as: (1) the ability of the metric value to describe the glare scale (evaluated by Spearman rank correlation), and (2) the ability of the metric to distinguish between disturbing and non-disturbing situations (evaluated by diagnostic receiver operating characteristic curve analysis tests). Furthermore, we defined robustness as the ability of a metric to deliver meaningful results when applied to different datasets and to fail as few as possible statistical tests. Average Spearman rank correlation coefficients in the range of 0.55–0.60 as well as average prediction rates to distinguish between disturbing and non-disturbing glare of 70–75% for several of the metrics indicate their reliability. The results also show that metrics considering the saturation effect as a main input in their equation perform better and are more robust in daylight-dominated workplaces than purely contrast-based metrics or purely empirical metrics. In this study, the daylight glare probability (DGP) delivered the highest performance amongst the tested metrics and was also found to be the most robust. Future research should aim to optimise the terms of glare equations which combine contrast and saturation effects, such as DGP, PGSV or UGRexp, to achieve metrics that also perform reliably in dimmer lighting conditions than the ones explored in this study.

Discomfort glare prediction by using Unified Glare Rating

ABSTRACTGlare which affects comfort or causes distraction is known as discomfort glare and glare which affects the ability to see any object is known as disability glare. Discomfort glare is mainly caused by artificial lighting of the workplace. Discomfort glare can be measured by using Visual Comfort Probability (VCP) or Unified Glare Rating (UGR). This paper discuses the discomfort evaluation methods and presents the calculations of UGR of a specific workplace. The discomfort glare is calculated by using DIALux lighting simulation software, Techno Team LMK measuring system and by using developed Python program model. The results are compared and verified with the feedback received from occupants/users of that workspace. A comparison of the UGR values determined by the DIALux lighting simulation software, LMK measurement system and subjective estimation shows that the values obtained by LMK measurement system are more accurate and are matching with the feedback received from occupants of the workspace.

Evaluation of a hybrid photo-radiometer sky model compared with the Perez sky model

Daylight simulation results greatly rely on the accuracy of sky models. When generic CIE sky and Perez all-weather sky simulate daylighting scenes they simplify cloud distributions. The lack of cloud distributions and subtle luminance variations causes simulation result deviations. High Dynamic Range (HDR) image techniques provide a method of generating accurate and actual skies. This paper explores the accuracy of a horizontal hybrid phadiometer (HPR) sky model, which combines modeled physical descriptions of the sun and HDR sky images. The aim was to build on and enrich the findings of a previous study (Humann and McNeil, 2017).In this study, the HDR images within four office spaces were taken. Simultaneously, global horizontal illuminance and HDR sky images were recorded outside. Simulated luminance maps were generated under both the HPR and Perez skies. Daylight Glare Probability and vertical eye illuminance (Ev) were calculated from 112 groups of luminance maps. The simulated Ev under HPR and Perez skies presented relative RMSE of 21.3% and 23.7%, respectively. The frequencies of accurate glare prediction under HPR and Perez skies were 95.5% and 93.9%, respectively. The results indicated that both the horizontal HPR sky and the Perez sky simulate equally accurate luminance maps.

Daylight Discomfort Glare Evaluation with Evalglare: Influence of Parameters and Methods on the Accuracy of Discomfort Glare Prediction

Nowadays, discomfort glare indices are frequently calculated by using evalglare. Due to the lack of knowledge on the implications of the methods and parameters of evalglare, the default settings are often used. But wrong parameter settings can lead to inappropriate glare source detection and therefore to invalid glare indices calculations and erroneous glare classifications. For that reason, this study aims to assess the influence of several glare source detection methods and parameters on the accuracy of discomfort glare prediction for daylight. This analysis uses two datasets, representative of the two types of discomfort glare: saturation and contrast glare. By computing three different statistical indicators to describe the accuracy of discomfort glare prediction, 63 different settings are compared. The results suggest that the choice of an evalglare method should be done when considering the type of glare that is most likely to occur in the visual scene: the task area method should be preferred for contrast glare scenes, and the threshold method for saturation glare scenes. The parameters that should be favored or avoided are also discussed, although a deeper understanding of the discomfort glare mechanism and a clear definition of a glare source would be necessary to reliably interpret these results.

Annual daylight glare evaluation: Impact of weather file selection

This paper illustrates the results of daylight glare simulations of a simple office carried out using five different five weather data files (Chinese typical year weather, CSWD(C), CSWD(A), international weather for energy calculations and Meteonorm), for three Chinese locations (Beijing, Shanghai and Guangzhou). The aim is to analyse the impact of choosing one weather file or the other on the prediction of daylight glare in the interior environment. The results show that using the Chinese typical year weather and CSWD(C) files can make a better prediction of the time distribution of daylight glare than the others. Analysis of the total amount of daylight glare leads to the conclusion that the Chinese typical year weather file should be recommended when considering extreme weather conditions and the CSWD(C) or Meteonorm files should be recommended in normal weather conditions.