Islands of misfit buildings: Detecting uncharacteristic electricity use behavior using load shape clustering

Quintana, Matias; Arjunan, Pandarasamy; Miller, Clayton

doi:10.1007/s12273-020-0626-1

Cited by 25 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Processing of data for calibration and the use in the simulation process could be improved [60,61,62]. The influence of predicting localized behavior better could better influence the classification of building use types [63].…”

Section: Potential Impact On Building Analysis and Control Applicationsmentioning

confidence: 99%

Using Google Trends as a proxy for occupant behavior to predict building energy consumption

Fu,

Miller

2021

Preprint

Self Cite

View full text Add to dashboard Cite

In recent years, the availability of larger amounts of energy data and advanced machine learning algorithms has created a surge in building energy prediction research. However, one of the variables in energy prediction models, occupant behavior, is crucial for prediction performance but hard-to-measure or time-consuming to collect from each building. This study proposes an approach that utilizes the search volume of topics (e.g., education or Microsoft Excel) on the Google Trends platform as a proxy of occupant behavior and use of buildings. Linear correlations were first examined to explore the relationship between energy meter data and Google Trends search terms to infer building occupancy. Prediction errors before and after the inclusion of the trends of these terms were compared and analyzed based on the ASHRAE Great Energy Predictor III (GEPIII) competition dataset. The results show that highly correlated Google Trends data can effectively reduce the overall RMSLE error for a subset of the buildings to the level of the GEPIII competition's top five winning teams' performance. In particular, the RMSLE error reduction during public holidays and days with site-specific schedules are respectively reduced by 20-30% and 2-5%. These results show the potential of using Google Trends to improve energy prediction for a portion of the building stock by automatically identifying site-specific and holiday schedules.

show abstract

Section: Potential Impact On Building Analysis and Control Applicationsmentioning

confidence: 99%

Using Google Trends as a proxy for occupant behavior to predict building energy consumption

Fu,

Miller

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the literature, there are only a few instances of training data using occupants counts [52]. Previous work has shown that the way occupants use buildings is different than expected [53,54] and data from Wifi signals of occupants [55,56], Bluetooth localization [57,58,59], and even text mining [60] can capture signals that could be used as inputs in the modeling process. This type of modeling approach may also influence occupant-centric controls [61].…”

Section: Using Occupant-centric Operations Data In Building Energy Pr...mentioning

confidence: 99%

Limitations of machine learning for building energy prediction: ASHRAE Great Energy Predictor III Kaggle competition error analysis

Miller,

Picchetti,

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning for building energy prediction has exploded in popularity in recent years, yet understanding its limitations and potential for improvement are lacking. The ASHRAE Great Energy Predictor III (GEPIII) Kaggle competition was the largest building energy meter machine learning competition ever held with 4,370 participants who submitted 39,403 predictions. The test data set included two years of hourly electricity, hot water, chilled water, and steam readings from 2,380 meters in 1,448 buildings at 16 locations. This paper analyzes the various sources and types of residual model error from an aggregation of the competition's top 50 solutions. This analysis reveals the limitations for machine learning using the standard model inputs of historical meter, weather, and basic building metadata. The types of error are classified according to the amount of time errors occur in each instance, abrupt versus gradual behavior, the magnitude of error, and whether the error existed on single buildings or several buildings at once from a single location. The results show machine learning models have errors within a range of acceptability (RMSLE scaled =< 0.1) on 79.1% of the test data. Lower magnitude model errors (0.1 < RMSLE scaled =< 0.3) occur in 16.1% of the test data. These discrepancies can likely be addressed through additional training data sources or innovations in machine learning. Higher magnitude errors (RMSLE scaled > 0.3) occur in 4.8% of the test data and are unlikely to be accurately predicted regardless of innovation. There is a diversity of error behavior depending on the energy meter type (electricity prediction models have unacceptable error in under 10% of test data, while hot water is over 60%) and building use type (public service less than 14%, while technology/science is just over 46%). Most of the instances of continuous error last longer than three days, and significant portions come from individual buildings and collections of buildings from the same site. This analysis forms the foundation for suggestions to reduce machine learning errors by collecting and using additional training data from onsite and web-based sources to improve the capability, accuracy, scalability, and usability of machine learning. Improvement of these metrics could enhance greater adoption in built environment applications.

show abstract

“…Among these aspects, occupancy schedule and appliance use schedule are two essential inputs requiring data-driven analysis of the typical profiles. Pattern identification is a major approach to understanding the occupant-behavior-related energy profiles in buildings (Yang et al 2018a;Wen et al 2019;Quintana et al 2020;Popoola and Chipango 2020), and is widely applied in both residential and non-residential buildings. This provides insight for distinguishing different types of users or (Sala et al 2019;Escobar et al 2020).…”

Section: Parametric Analysis At the Design Phasementioning

confidence: 99%

Advanced data analytics for enhancing building performances: From data-driven to big data-driven approaches

et al. 2020

View full text Add to dashboard Cite

Buildings have a significant impact on global sustainability. During the past decades, a wide variety of studies have been conducted throughout the building lifecycle for improving the building performance. Data-driven approach has been widely adopted owing to less detailed building information required and high computational efficiency for online applications. Recent advances in information technologies and data science have enabled convenient access, storage, and analysis of massive on-site measurements, bringing about a new big-data-driven research paradigm. This paper presents a critical review of data-driven methods, particularly those methods based on larger datasets, for building energy modeling and their practical applications for improving building performances. This paper is organized based on the four essential phases of big-data-driven modeling, i.e., data preprocessing, model development, knowledge post-processing, and practical applications throughout the building lifecycle. Typical data analysis and application methods have been summarized and compared at each stage, based upon which in-depth discussions and future research directions have been presented. This review demonstrates that the insights obtained from big building data can be extremely helpful for enriching the existing knowledge repository regarding building energy modeling. Furthermore, considering the ever-increasing development of smart buildings and IoT-driven smart cities, the big data-driven research paradigm will become an essential supplement to existing scientific research methods in the building sector.

show abstract

Islands of misfit buildings: Detecting uncharacteristic electricity use behavior using load shape clustering

Cited by 25 publications

References 32 publications

Using Google Trends as a proxy for occupant behavior to predict building energy consumption

Using Google Trends as a proxy for occupant behavior to predict building energy consumption

Limitations of machine learning for building energy prediction: ASHRAE Great Energy Predictor III Kaggle competition error analysis

Advanced data analytics for enhancing building performances: From data-driven to big data-driven approaches

Contact Info

Product

Resources

About