Search NASA⌕ Search

SEARCH · Search NASA

Results for “Linear Regression”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 451 records · Page 25

Performance of low-cost indoor air quality monitors for PM 2.5 and PM 10 from residential sources

Advances in particle sensor design and manufacturing have enabled the development of low-cost air quality monitors (LCMs). Here, the sensors use light scattering to estimate mass concentration and thus require evaluation for aerosols of varied composition and size distribution. We tested the performance of six LCMs designed for home use and having a retail price under US$300 in October 2018. We assessed their performance by comparing their output to reference PM 2.5 and PM 10 measurements from 21 common residential sources and from infiltrated outdoor PM 2.5 . Reference data were obtained by using gravimetric measurements to adjust time-resolved output from an aerosol spectrometer with both electrical mobility and optical particle sensors. Compared by linear regression to reference measurements, LCMs had negative intercepts and slopes of 1–2 for infiltrated outdoor PM 2.5 . Semi-quantitative responses (~50–200% of actual PM 2.5 ) were obtained for varied aerosols including minerals (ultrasonic humidifier, vacuuming, test dust); combustion products (incense, mosquito coil, extinguished candles); microwave popcorn; and cooking involving frying or grilling. LCMs had low or no response to sources for which all mass was in particles smaller than 0.25 μm, including steady candle flames and cooking without frying or grilling. PM 10 data from LCMs was more variable than PM 2.5 .

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

The impacts of building characteristics, social psychological and cultural factors on indoor environment quality productivity belief

Indoor environmental quality (IEQ) plays a key role in determining occupants' productivity at work; however, analyses of the interconnected factors among building physical, attitudinal, social and demographic components in one study are lacking. To fill this research gap, this study investigates these interconnected factors' influence on occupants’ IEQ-productivity belief, defined as a personal, subjective evaluation of the linkage between the impacts of five IEQ aspects (the quality of indoor temperature, air, natural and electric lighting, and acoustics) and productivity. A cross-sectional survey data was collected in university offices from six countries (Brazil, Italy, Poland, Switzerland, Taiwan and the U.S.). Results of multiple linear regression models indicate that IEQ satisfaction is the strongest positive predictor of the IEQ-productivity belief and this relationship is stronger in private offices. Country of residence is the second primary predictor. Several attitudinal-behavioral factors, including thermal comfort, perceived ease of controlling indoor environmental features, and attitudes toward sharing controls are all positively associated with IEQ-productivity belief. Interestingly, the level of control accessibility to light switches has the strongest impact as opposed to other controls. On the other hand, group norms and conformity intention are not significant predictors. Regarding demographics, men are more likely than women to perceive the IEQs to have positive impacts on their productivity, without considering other variables in the regression model; however, women are more likely than men to consider all IEQs as having positive impacts on productivity, after considering other variables. Furthermore, our findings provide suggestions for prioritizing wellness in the workplace since the early design stage.

99 GENERAL AND MISCELLANEOUS↗

Machine learning–assisted prediction of heat fluxes through thermally anisotropic building envelopes

Thermally anisotropic building envelope (TABE) is a novel active building envelope that can save energy use to maintain thermal comfort in buildings by redirecting heat and coolness from building envelopes to thermal loops. Finite element models (FEMs) can be used to compute the heat fluxes through TABEs, but the high computational cost of finite element simulations has prevented parametric studies and design optimizations. This paper proposes a domain knowledge–informed, finite element–based machine learning framework to reduce the computation cost for the energy management of buildings installed with TABE that uses a ground thermal loop. First, the training heat flux data set was generated by FEM simulations with different thermal loop schedules. Then, both shallow learning models (i.e., multivariate linear regression and eXtreme Gradient Boost, or XGBoost) and a deep learning model (i.e., deep neural network, or DNN) were trained to predict the heat fluxes. Domain knowledge was used for data preprocessing and feature selection. Finally, the suitability of the selected machine learning model was tested under different thermal loop schedules. Herein, the case study results showed that: (1) XGBoost can be as accurate as DNN (coefficient of determination equal to 0.81) with much less training time; (2) the annual energy cost savings for different thermal loop schedules obtained by the XGBoost-predicted and FEM-calculated heat fluxes are consistent, having a difference of only 4%; and (3) XGBoost can reduce the computation time for the annual energy analysis of the case study building with a given thermal loop schedule from around 12 h by using FEM to less than 1 min.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

The role of building characteristics, demographics, and urban heat islands in shaping residential energy use

This study assesses the effects of residential building structure, socio-demographics, and ambient temperature conditions of a neighborhood to overall energy expenditures of a household. Using annual utility billing data, existing tax-lot records, temperature data, and U.S. Census demographic data, we ask two research questions: (1) What role do differences in neighborhood ambient heat, structural conditions and behavior play in helping to explain residential energy use? And (2) In what ways do energy expenditures spatially cluster in relation to urban heat islands? Using linear regression, spatial regression, and spatial clustering techniques, we evaluated the role of physical location of these households in relation to ambient temperatures. Significant positive relationships among energy expenditures and building size, average household size, White population, and population over 65 were observed; Significant negative relationships among energy expenditures and building density, renter population and population with a bachelor’s degree or higher were also observed. Negative relationships among morning and afternoon measures of heat, suggest that some areas spend less on energy use as a result of the neighborhood scale thermal environment. Spatial analysis of ambient temperature and energy expenditures show that energy use throughout the city is spatially clustered, with more affluent areas using more energy, regardless of whether heat islands are present. The results help to frame the role that increasing ambient temperatures in cities play in the spatial disparity in energy demand.

Residential Buildings, Energy Expenditure, Socio-d↗

Experimental determination of the thermodynamic parameters affecting the adsorption behaviour and dispersion effectiveness of PCE superplasticizers

For adsorption of three different allylether-based PCE superplasticizers on CaCO{sub 3} surface, the thermodynamic parameters DELTAH, DELTAS and DELTAG were determined experimentally. The GIBBS standard free energy of adsorption DELTAG{sub 0ads}, the standard enthalpy of adsorption DELTAH{sub 0ads} and the standard entropy of adsorption DELTAS{sub 0ads} applying to an unoccupied CaCO{sub 3} surface were obtained via a linear regression of ln K (equilibrium constant) versus 1 / T (VAN'T HOFF plot). Additionally, the thermodynamic parameters characteristic for a CaCO{sub 3} surface loaded already with polymer (isosteric conditions) were determined using a modified CLAUSIUS-CLAPEYRON equation. For all PCE molecules, negative DELTAG values were found, indicating that adsorption of these polymers is energetically favourable and a spontaneous process. Adsorption of PCEs possessing short side chains is mainly instigated by electrostatic attraction and a release of enthalpy. Contrary to this, adsorption of PCEs with long side chains occurs because of a huge gain in entropy. The gain in entropy results from the release of counter ions attached to the carboxylate groups of the polymer backbone and of water molecules and ions adsorbed on the CaCO{sub 3} surface. With increased surface loading, however, DELTAG{sub isosteric} decreases and adsorption ceases when DELTAG becomes 0. The presence of Ca{sup 2+} ions in the pore solution strongly impacts PCE adsorption, due to complexation of carboxylate groups and a reduced anionic charge amount of the molecule. In the presence of Ca{sup 2+}, adsorption of allylether-based PCEs is almost exclusively driven by a gain in entropy. Consequently, PCEs should produce a strong entropic effect upon adsorption to be effective cement dispersants. Molecular architecture, anionic charge density and molecular weight as well as the type of anchor groups present in a superplasticizer determine whether enthalpy or entropy is the dominant force for superplasticizer adsorption.

36 MATERIALS SCIENCE↗

Can Zn isotopes in sediments record past eutrophication of freshwater lakes? A pilot study at Lake Baldegg (Switzerland)

In this study, the speciation and isotopic composition of Zn were traced across the sediments of a freshwater lake that experienced one hundred years of strong eutrophication, in order to assess the potential of sedimentary Zn isotopes to record such an environmental disturbance. Here, the results indicate that the sedimentary Zn isotope signal varied with the change from pre-eutrophic to eutrophic conditions in the investigated lake. The average δ 66 Zn JMC value of the dominantly allochthonous lithogenic sediments deposited during the pre-eutrophic period was +0.27‰ ±0.05‰ (i.e. similar to the δ 66 Zn JMC value of +0.28‰ ±0.05‰ proposed for Bulk Silicate Earth), while enhanced autochthonous biochemical sedimentation during the eutrophic period resulted in significantly lower δ 66 Zn JMC values down to +0.04‰ ±0.06‰. Synchrotron-based X-ray absorption spectroscopy data revealed a concomitant change in Zn speciation from a dominant fraction of Zn in clay minerals during the pre-eutrophic period to a major fraction of Zn in ZnS during the eutrophic period. A linear regression relating the sedimentary Zn isotope signal to the fraction of Zn in ZnS indicated δ 66 Zn JMC values of +0.27‰ ±0.06‰ and 0.00‰ ±0.08‰ for Zn in clay minerals and in ZnS, respectively. The enrichment of light Zn in ZnS in the eutrophic sediments is tentatively attributed to enhanced biological uptake of light Zn in the water column, which resulted in an enhanced flux of organic-bound Zn towards the sediments and further transformation of organic Zn into ZnS upon biomass mineralization during early diagenesis. This hypothesis is in agreement with the fractionation towards lighter Zn reported for both biological uptake of Zn and ZnS precipitation. The results of this study emphasize the potential of sedimentary Zn isotopes to register past eutrophic periods in freshwater lakes, and thus to serve as a probe of paleo-environmental conditions and/or past land use at the catchment scale.

58 GEOSCIENCES↗

Reinforced double-threaded slide-ring networks for accelerated hydrogel discovery and 3D printing

Traditionally, slide-ring gels are stretchable but soft as a result of an elasticity-stretchability trade-off. Herein, we introduce a new approach to breaking this trade-off and creating reinforced slide-ring networks with mobile crosslinkers. Our approach involves the construction of a polyethylene glycol double-threaded γ-cyclodextrin-based pro-slide-ring crosslinker that serves as a modular component for 3D printing and copolymerization. The resulting crystalline-domain-reinforced slide-ring hydrogels, or CrysDoS-gels, exhibit both high elasticity and high stretchability. The modular synthesis allows for high-throughput synthesis of CrysDoS-gels, generating a large amount of data for structure-property analysis. Here, by employing data science techniques, such as machine learning and linear regression, not only were we able to identify which chemical components influence the mechanical properties of CrysDoS-gels, but this analysis also aided in the discovery of better-performing CrysDoS-gels. Finally, we demonstrate the potential application of the newly discovered CrysDoS-gels as sensing devices by 3D printing them as stress sensors with high sensitivity and a broad detection range.

3D-printing↗

Derivative-based SINDy (DSINDy): Addressing the challenge of discovering governing equations from noisy data

Recent advances in the field of data-driven dynamics allow for the discovery of ODE systems using state measurements. One approach, known as Sparse Identification of Nonlinear Dynamics (SINDy), assumes the dynamics are sparse within a predetermined basis in the states and finds the expansion coefficients through linear regression with sparsity constraints. This approach requires an accurate estimation of the state time derivatives, which is not necessarily possible in the high-noise regime without additional constraints. We present an approach called Derivative-based SINDy (DSINDy) that combines two novel methods to improve ODE recovery at high-noise levels. First, we denoise the state variables by applying a projection operator that leverages the assumed basis for the system dynamics. Second, we use a second order cone program (SOCP) to find the derivative and governing equations simultaneously. We derive theoretical results for the projection-based denoising step, which allow us to estimate the values of hyperparameters used in the SOCP formulation. This underlying theory helps limit the number of required user-specified parameters. Finally, we present results demonstrating that our approach leads to improved system recovery for the Van der Pol oscillator, the Duffing oscillator, the Rössler attractor, and the Lorenz 96 model.

97 MATHEMATICS AND COMPUTING↗

Oxidation states of binary oxides from data analytics of the electronic structure

Despite their simplicity, the concepts of oxidation states (OS) and Bader charge are widely used in the analysis of experimental data and theoretical calculations for gaining physical and chemical insights. However, a connection between the OS and Bader charge has not been generally demonstrated. To our knowledge, all previous works concerning this connection were limitted to a small number of compounds. The aim of this work was to find a dependency between OS and Bader charge for a large number of compounds from an open database. Using a simple machine learning algorithm, linear regression, we show that a correlation indeed exists for the binary oxide systems. The applicability of determining OS by Bader charges for binary oxides in mixed-valence compounds and surfaces is considered and the limits of validity of the approach is discussed.

36 MATERIALS SCIENCE↗

Machine learning elastic constants of multi-component alloys

The present manuscript explores application of machine learning methods for determining elastic constants and other derived mechanical properties of multi-component alloys. Here, a number of machine learning models, including linear regression, neural network and random forest based models, are trained and tested on a dataset of binary alloys generated using density functional theory (DFT) calculations and spanning over a large number of elemental species in the periodic table. Starting with a wide range of simple and easily accessible compositionally-averaged elemental features, a correlation-based feature selection strategy was used to systematically down-select a set of most relevant features towards the prediction of the elasticity tensor components. The true predictive performance and the associated uncertainties of the models were established by testing on unseen data and bootstrapping, respectively. A single and pair-wise feature partial dependence analysis was performed to visualize the average property trends in the multi-dimensional feature space in order to further understand the achieved predictive performance. The utility of the trained model is further demonstrated by obtaining sufficiently accurate yet highly efficient approximations for bulk modulus, Young’s modulus, shear modulus and Poisson’s ratio for alloys beyond the binary space (i.e., two-component alloys) on which the model was originally trained. More importantly, we test and validate the predictive performance of the developed model directly against the experimentally measured elastic constants of technologically relevant multi-component alloys (such as, Ni- and Ti-based alloys). Finally, utility of such a data-enabled route is demonstrated by predicting the possible range of various elastic properties for vast composition space available within the five component Ni-Cr-Fe-Mo-W alloy system in a high-throughput manner.

36 MATERIALS SCIENCE↗

Machine learning-enabled prediction of chemical durability of A 2 B 2 O 7 pyrochlore and fluorite

Pyrochlore-structure type and its derivative in a general formula A 2 B 2 O 7 (A = rare earth elements and actinides; B = Ti, Sn, Zr, Hf, Pb, Si, etc.) display excellent structural flexibility and rich crystal chemistry as promising nuclear waste form materials capable of immobilizing actinides and fission products. It is essential to understand these materials’ chemical durability and element release of radionuclides in order to evaluate their performance in near-field environment. However, it is a formidable grand technological challenge to experimentally perform durability testing across hundreds of thousands of possibilities resulting from their extreme compositional complexities due to cation substitutions at both A and B-sites. In this work, we demonstrate a machine learning approach to determine the key materials parameters and structural characteristics governing the leaching behaviors from a small set of selected compositions as model systems, enabling a science-based prediction of their chemical durability that can be extended to a wide range of chemical compositions. The combination of four key structural characteristics and materials parameters, including ionic radius size difference , ionic potential difference , electronegativity difference , and lattice parameter , creates features an optimized prediction of the chemical durability. Two machine learning models, linear regression and Kernel ridge regression models, are trained on the randomly-split training dataset derived from the experimentally-determined elemental release rates, and subsequently tested on the testing dataset. The predicted leaching rates from both machine learning models show an excellent agreement with the experimental data, demonstrating the feasibility of rapidly evaluating the material properties of new compositions. These results highlight the immense potential of synergizing informatics through machine learning-based models and well-controlled experiments of selected model systems to accelerate materials design and discovery with optimized compositions and performance of promising materials for effective nuclear waste management.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Optical image analysis for graphene layer detection: Enhanced green channel methodology

Graphene, a material of increasing research interest, requires accurate layer identification due to its sensitivity to layer count. Existing methods for graphene layer number identification are either time-consuming or of low accuracy, with high-accuracy methods often requiring expensive processes. This paper aims to address this challenge by proposing a cost-effective and efficient approach. Specifically, the current work highlights only the green channel—one of the three primary color channels (red, green, blue) that make up an optical image—from images of exfoliated graphene flakes for layer count identification. A linear regression is performed between pixel position and substrate green channel value, and this effect is subtracted from the entire optical image to mitigate background effects. By storing the range of green channel values for each type of flake (monolayer, bilayer, or tri-layer) based on a few images, we establish thresholds for identifying different types of layers in a particular setup. Additionally, our methodology allows for flexible threshold tuning using a single reference image, enabling adjustment to changes in detection setup such as illumination level, magnification, or microscope used. Finally, demonstrating high accuracy and flexibility, this methodology presents a suitable technique for graphene layer number identification without the need for large datasets or expensive instruments.

2D materials↗

Biomarkers of Tuberculosis Severity and Treatment Effect: A Directed Screen of 70 Host Markers in a Randomized Clinical Trial

More efficacious treatment regimens are needed for tuberculosis, however, drug development is impeded by a lack of reliable biomarkers of disease severity and of treatment effect. In this study, we conducted a directed screen of host biomarkers in participants enrolled in a tuberculosis clinical trial to address this need. Serum samples from 319 protocol-correct, culture-confirmed pulmonary tuberculosis patients treated under direct observation as part of an international, phase 2 trial were screened for 70 markers of infection, inflammation, and metabolism. Biomarker assays were specifically developed for this study and quantified using a novel, multiplexed electrochemiluminescence assay. We evaluated the association of biomarkers with baseline characteristics, as well as with detailed microbiologic data, using Bonferroni-adjusted, linear regression models. Across numerous analyses, seven proteins, SAA1, PCT, IL-1, IL-6, CRP, PTX-3 and MMP-8, showed recurring strong associations with markers of baseline disease severity, smear grade and cavitation; were strongly modulated by tuberculosis treatment; and had responses that were greater for patients who culture-converted at 8 weeks. With treatment, all proteins decreased, except for osteocalcin, MCP-1 and MCP-4, which significantly increased. Several previously reported putative tuberculosis-associated biomarkers (HOMX1, neopterin, and cathelicidin) were not significantly associated with treatment response. In conclusion, across a geographically diverse and large population of tuberculosis patients enrolled in a clinical trial, several previously reported putative biomarkers were not significantly associated with treatment response, however, seven proteins had recurring strong associations with baseline radiographic and microbiologic measures of disease severity, as well as with early treatment response, deserving additional study.

60 APPLIED LIFE SCIENCES↗

Does size matter? An evaluation of length and proportion of information in environmental impact statements

Highlights: • Empirical analysis of length and proportion of information in EISs is presented. • Current EISs, in Brazil, are significantly longer than the early ones. • Baseline information is the most common element in EISs. • EIS length is affected by many variables, including project sector and size. • While relevant to decision-making, EIS length should be carefully regulated. - Abstract: For decades, authors and institutions have argued that the quality of Environmental Impact Statements (EISs) is somehow affected by the volume of information they contain. Both too little and too much information can be a problem. However, very few academic studies have addressed the issue of EIS length in detail. The objective of this article was to systematically analyze the volume of information presented in EISs, using Brazil as the empirical context. More specifically, this study evaluated the volume and proportion of information disclosed in 49 Brazilian EISs. This study also tried to identify sectorial variations and whether variables such as project size and number of pages in Terms of References are likely determinants of information volume. >146 thousand pages of EIS information were scrutinized in two rounds of content analysis. Data were organized in spreadsheets and then coded and analyzed through various descriptive and inferential statistical techniques. Overall, findings corroborate the fact that EISs are now significantly longer than the early ones, and still heavily loaded with baseline information. The average number of pages in EISs and in Non-technical Summaries was found to be 2993 and 94, respectively. Kruskal-Wallis and linear regression tests indicated that EIS length is likely affected by a combination of variables, including project size, territorial and sectorial characteristics. Such findings suggest that the historical approach of setting page limits to EISs through regulations and Terms of References is no longer appropriate for EIA practice in connection with large enterprises in Brazil, and arguably elsewhere. The article discusses its practical and academic implications, and highlights the need to further investigate the actual impacts of EIS length on decision-making.

54 ENVIRONMENTAL SCIENCES↗

Gradient boosting machine for modeling the energy consumption of commercial buildings

Accurate savings estimations are important to promote energy efficiency projects and demonstrate their cost-effectiveness. The increasing presence of advanced metering infrastructure (AMI) in commercial buildings has resulted in a rising availability of high frequency interval data. These data can be used for a variety of energy efficiency applications such as demand response, fault detection and diagnosis, and heating, ventilation, and air conditioning (HVAC) optimization. This large amount of data has also opened the door to the use of advanced statistical learning models, which hold promise for providing accurate building baseline energy consumption predictions, and thus accurate saving estimations. The gradient boosting machine is a powerful machine learning algorithm that is gaining considerable traction in a wide range of data driven applications, such as ecology, computer vision, and biology. In the present work an energy consumption baseline modeling method based on a gradient boosting machine was proposed. To assess the performance of this method, a recently published testing procedure was used on a large dataset of 410 commercial buildings. The model training periods were varied and several prediction accuracy metrics were used to evaluate the model's performance. The results show that using the gradient boosting machine model improved the R-squared prediction accuracy and the CV(RMSE) in more than 80 percent of the cases, when compared to an industry best practice model that is based on piecewise linear regression, and to a random forest algorithm.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

A cross-dimensional analysis of data-driven short-term load forecasting methods with large-scale smart meter data

Electricity load forecasting is essential to utility operation and power grid stability. A wide spectrum of data-driven methods, ranging from linear regression models to more recent deep learning models have been adopted to forecast electric load over the years. However, there still lacks a holistic evaluation of the applicability of conventional statistical and machine learning based algorithms with respect to different temporal and spatial scopes, computational requirements, and sensitivity of model-tuning. Enabled by a large-scale electricity load profile dataset of over 40,000 residential customers in a utility region, we conducted a cross-dimensional analysis of data-driven load forecasting methods. Three regression-based and seven deep learning algorithms with different model configurations were evaluated in terms of their overall and peak load prediction accuracy, and training burdens, across spatial aggregation levels ranging from the transformer, feeder, substation, to neighborhood. We found, first, the load forecasting accuracy is constrained by a predictability boundary, influenced by the forecasting horizon and spatial aggregation level. Specifically, RandomForest, XGBoost, TFT, TSMixer, and TiDE models achieved less than 10 % prediction error for up to 96-h ahead forecasting for district, substation, and feeder levels, while other models struggle at long-horizon predictions; Second, for winter and summer peak load dates, most models were able to predict the peak demand timing within ± 1 h, but the prediction percentage error varied by models, with TFT and TiDE models being the top performers; Third, models with similar prediction accuracy can differ in training burden by an order of magnitude. Therefore, choosing model configurations that balance prediction performance and computational resource is an important practical consideration for large-scale deployment of the machine learning based load forecasting. The outcome of this study can guide researchers and practitioners to choose the proper load forecasting algorithms based on their problem scope, required accuracy, and available resources. The predictability boundary can serve as a benchmark for electricity load forecasting problems with new algorithms and datasets.

Li, Han↗

Surrogate model evaluation and building energy benchmarking for commercial buildings

Building energy consumption benchmarking involves challenges associated with various energy patterns for different building types; heating, ventilating, and air-conditioning (HVAC) system types; and climates. Given significant variation in energy use patterns, accurate prediction of long-term energy use using surrogate models remains challenging. Multiple linear regression (MLR) is commonly used for building energy benchmarking because of its simple structure; however, it lacks accuracy compared to other black-box models. Although many studies have compared surrogate models and offer guidance on model selection based on metrics, they do not provide detailed analysis on improving the surrogate model accuracy. In this paper, we implement a surrogate model using polynomial ridge regression (i.e., MLR with interaction terms combined with ridge regularization) for small office and retail strip mall buildings across six HVAC system types and all climate zones, for electricity and natural gas in baseline and proposed scenarios. A simulation workflow is developed using OpenStudio TM /EnergyPlus TM to generate simulation data using measures over a wide range of efficiency inputs. Enhancements based on statistical insights are used for improving the model accuracy using filters, input transformations, and change points. Surrogate models achieved average coefficient of variation of the root mean squared error (CVRMSE) values of 2.17, 1.06, 2.05, and 3.26 for proposed electricity, proposed natural gas, baseline electricity, and baseline natural gas, respectively, with enhancements reducing CVRMSE by an average of 14.9% across all combinations. We provide model interpretation via Shapley additive explanations to determine which input variables most influence energy consumption and provide supportive arguments for enhancements.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Predicting plug loads with occupant count data through a deep learning approach

Predictive control has gained increasing attention for its ability to reduce energy consumption and improve occupant comfort in buildings. The plug loads prediction is a key component for the predictive building controls, as plug loads is a major source of internal heat gains in buildings. This study proposed a novel method to apply the Long-Short-Term-Memory (LSTM) Network, a special form of Recurrent Neural Network, to predict plug loads. The occupant count and the time have been confirmed to drive the plug load profile and thus selected as the features for the plug load prediction. The LSTM network was trained and tested with ground truth occupant count data collected from a real office building in Berkeley, California. Results from the LSTM network markedly improve the prediction accuracy compared with traditional linear regression methods and the classical Artificial Neural Network. 95% of 1-hour predictions from LSTM network are within ±1 kW of the actual plug loads, given the average plug loads during the office hour is 8.6 kW. The CV(RMSE) of the predicted plug load is 11% for the next hour, and 20% for the next eight hours. Lastly, we compared four prediction approaches with the office building we monitored: LSTM vs. ARIMA, with occupant counts vs. without occupant counts. It was found, the prediction error of the LSTM approach is around 4% less than the ARIMA approach. Using occupant counts as an exogenous input could further reduce the prediction error by 5%-6%. The findings of this paper could shed light on the plug load prediction for building control optimizations such as model-predictive control.

Occupant count↗