No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
User Behavior Modeling for Estimating Residential Energy Consumption
Abstract
Residential energy constitutes a significant portion of the total US energy consumption. Several researchers proposed energy-aware solutions for houses, promising significant energy and cost savings. However, it is important to evaluate the outcomes of these methods on larger scale, with hundreds of houses. This paper presents a human-activity based residential energy modeling framework, that can create power demand profiles considering the characteristics of household members. It constructs a mathematical model to show the detailed relationships between human activities and house power consumption. It can be used to create various house profiles with different energy demand characteristics in a reproducible manner. Comparison with real data shows that our model captures the power demand differences between different family types and accurately follows the trends seen in real data. We also show a case study that evaluates voltage deviation in a neighborhood, which requires accurate estimation of the trends in power consumption.
... Earlier researchers recognized that understanding human behaviors is an important task to accomplish such goals. For example, diverse techniques exploited human activities and contexts as key control knobs of various system managements including mobile systems [9] and smart homes [1]. ...
... For example, prior research has shown that understanding users behavior and exploiting the behavioral characteristics can be used to improve system efficiency. In this context, earlier work proposed diverse system optimization techniques by identifying user behaviors and interactions for mobile systems [9] and smart homes [1]. Prior work often utilized ML techniques to identify the activities, while relying on computing capability of clouds through offloading, e.g., [18]. ...
... The results are reported for the non-binarized models, since the overhead of the model binarization is negligible. 1 In this comparison, the HD modeling and the neural network training were both executed on x86 processor. The results show that the proposed method presents higher performance efficiency as compared to the neural network training. ...
Human activity recognition is a key task of many Internet of Things (IoT) applications to understand underlying contexts and react with the environments. Machine learning is widely exploited to identify the activities from sensor measurements, however, they are often overcomplex to run on less-powerful IoT devices. In this paper, we present an alternative approach to efficiently support the activity recognition tasks using brain-inspired hyperdimensional (HD) computing. We show how the HD computing method can be applied to the recognition problem in IoT systems while improving the accuracy and efficiency. In our evaluation conducted for three practical datasets, the proposed design achieves the speedup of the model training by up to 486x as compared to the state-of-the-art neural network training. In addition, our design improves the performance of the HD-based inference procedure by 7x on a low-power ARM processor.
... One of the popular survey techniques is the contingent valuation method (CVM) (Arrow et al., 1993), which is based on directly inquiring the household's willingness-to-pay (WTP) or willingness-to-accept (WTA) to explore the marginal utility from a change in consumption of specific non-market goods. More and more researches became interested in the residential energy consumption behavior, such as the behavior-lead energy consumption patterns (Zhou and Yang, 2016;Aksanli et al., 2016) and rebound effect (Azevedo, 2014), which could be an important influencing factor to the DR programs. ...
This work is a combined study of the economics-based customer survey and demand response (DR) optimization. We present a methodology to capture the customers’ behavior and thermal parameter differences within the decision-making of the voluntary air-conditioning DR program for critical peak reduction. Firstly, we developed a face-to-face survey based on the contingent valuation method (CVM) to investigate the demand-side bid preferences of residential customers, while the households’ willingness-to-accept (WTA) distribution on the particular AC interruptions is evaluated by an expenditure difference model (EDM). Based on the survey, we decoupled the inner relationships between the customers’ bidding behavior versus consumption properties through an aggregate thermal state analysis of the customers’ entity. Subsequently, the above data are fitted into an optimization framework with DR modeling to maximize the system’s benefit, optimize the DR control, and find the optimal reward price on saving the system’s generation investment with critical peak reduction. In the DR modeling, we presented a group thermal state model, in which the customers are divided into several groups and respond to the DR signal sequentially to mitigate the behavior-lead cooling-rebound. Besides, we considered the responding convenience of the customers, which is based on their real-time stay-at-home rate and thermal state evaluation during the summer peak. The proposed methodology could strongly support the system’s decision-making and is finally validated within the city of Xi’an, China.
... However, their approach is for modelling the activities of the people who work in the industrial sector. In addition to the probability models, Aksanli et al. develop a graph-based model to represent the chain of user activities [18], and Basu et al. use a decision tree in their modelling [19]. Table 1 summarises the reviewed literature that use TUS data for activity modelling. ...
User activities is an important input to energy modelling, simulation and performance studies of residential buildings. However, it is often difficult to obtain detailed data on user activities and related energy consumption data. This paper presents a stochastic model based on Markov chain to simulate user activities of the households with one or more family members, and formalizes the simulation processes under different conditions. A data generator is implemented to create fine-grained activity sequences that require only a small sample of time-use survey data as a seed. This paper evaluates the data generator by comparing the generated synthetic data with real data, and comparing other related work. The results show the effectiveness of the proposed modelling approach and the efficiency of generating realistic residential user activities.
... However, there are also models that are not based on clustering techniques. For example, in [17] the authors establish a framework that is able to establish profiles of energy demand in residential areas by means of a mathematical model that details the relation- ship between human activity and energy consumption. In addition, use an autoregressive moving average model (ARMA) to detect malicious consumption patterns due to electrical intrusions [18]. ...
Thanks to the presence of sensors and the boom in technologies typical of the Internet of things, we can now monitor and record the energy consumption of buildings over time. By effectively analyzing these data to capture consumption patterns, significant reductions in consumption can be achieved and this can contribute to a building’s sustainability. In this work, we propose a framework from which we can define models that capture this casuistry, gathering a set of time series of electrical consumption. The objective of these models is to obtain a linguistic summary based on y is P protoforms that describes in natural language the consumption of a given building or group of buildings in a specific time period. The definition of these descriptions has been solved by means of fuzzy linguistic summaries. As a novelty in this field, we propose an extension that is able to capture situations where the membership of the fuzzy sets is not very marked, which obtains an enriched semantics. In addition, to support these models, the development of a software prototype has been carried out and a small applied study of actual consumption data from an educational organization based on the conclusions that can be drawn from the techniques that we have described, demonstrating its capabilities in summarizing consumption situations. Finally, it is intended that this work will be useful to managers of buildings or organizational managers because it will enable them to better understand consumptionin a brief and concise manner, allowing them to save costs derived from energy supply by establishing sustainable policies.
... Given buildings' overall significant contribution to energy use, as well as environmental concerns and climate change, methods are needed to reduce this consumption. This is particularly the case for residential buildings, whose operation is highly dependent on occupants and their behavior [3][4][5][6]. There are many possible strategies to reduce energy use in residential buildings; the most common of which is through retrofitting an existing building with more energy efficient systems. ...
Purpose of Review
Residential energy performance prediction has historically received less attention, as compared to commercial buildings. This likely is in part due to the limited availability of residential energy data, as well as the relative challenge of predicting energy consumption of buildings that are more highly dependent on occupant behavior. The purpose of this effort is to assess the types and characteristics of energy and non-energy data available for algorithm developed and methods that have been developed to predict residential consumption.
Recent Findings
While there are several large residential building energy datasets, data availability is still generally very limited. Most energy prediction methods used recently include data-driven approaches, as well as combinations of multiple methods; however, many methods have not been tested for residential buildings, or at a range of energy data frequencies.
Summary
The literature points to the need for the availability of more residential building data sources to be able to assess and improve models, and further testing is needed including those models that have not yet been significantly used for residential buildings.
... The methods proposed to solve the problem of noisy and missing load data can be broadly divided into three categories: machine learning approaches, mathematical and statistical methods [10,12,[26][27][28][29][30][31][32][33][34][35][36][37][38][39][40]. In this section, we offer a form of filtration for pre-processing the hourly load training dataset that is employed for the short-term load forecasting. ...
This paper offers a form of filtration based on moving average filter and KNN imputation method, for pre-processing hourly electricity load data for Short-Term Load Forecasting (STLF). The STLF is developed by the Adaptive Network Based Fuzzy Inference System (ANFIS). There is a lack of data pre-processing related to load forecasting, especially STLF. Unlike previous studies, to enhance the accuracy of forecasting, the current study considers data pre-processing as well. We propose a machine learning model using the ANFIS to forecast short-term load. The electricity load data are used for training and testing the proposed model. The predictor's outputs show that the model able to forecast electricity load in an accurate way. We believe the proposed pre-processing method can be used in the future studies to increase forecast accuracy.
... In future work, we will investigate a decentralized mechanism and consider direct cooperation among households forming small coalitions to reduce their joint peak demand further. We will model more complex user behaviors than at present and a variety of appliances [37]. In particular, we are interested in approaches that not only reduce peak demand but reduce aggregate demand (i.e., save power not just shift load). ...
Given the improved understanding of sustainability, hundreds of factors are identified to have relevance to building energy efficiency. However, there is still a lack of knowledge about what factors play a significant role in energy consumption prediction for residential buildings. In the absence of this information, building energy consumption prediction would not be efficient. To tackle this problem, this study creates a feature engineering-based analytic framework to select effective factors for energy consumption prediction and assess their implications. Two application cases are reported to demonstrate the efficiency improvement of energy consumption prediction for residential buildings. The cases use the Residential Energy Consumption Survey database that contains more than 270 energy use-related factors about buildings and occupants in the United States. Data analysis from the two cases shows that selected features achieve 97-102% of prediction power while using 12-15% number of factors, largely reducing the dimensionality for energy prediction. The results also produce a list of significant features that are efficient predictors for residential energy modeling and evaluation at the national and regional levels. Examples of the selected features are the total number of rooms and full bathrooms, frequency of clothes dryer used, type of the housing unit, number of ceiling fans and television. The selected features explain energy use patterns and their relationships which help designer, contractors, and occupants better understand energy, behaviors, and the built environment. The resultant energy use patterns inform regional similarities, differences, and distinctive characteristics.
Energy Management Systems (EMS) are mainly price driven with minimal consumer interaction. To improve the effectiveness of EMS in the context of demand response, an alternative EMS control framework driven by resident behavior patterns is developed. Using hidden Markov modeling techniques, the EMS detects consumer behavior from real-time aggregate consumption and a pre-built dictionary of reference models. These models capture variations in consumer habits as a function of daily living activity sequence. Following a training period, the system identifies the best fit model which is used to estimate the current state of the resident. When a request to activate a time-shiftable appliance is made, the control agent compares grid signals, user convenience constraints, and the current consumer state estimate to predict the likelihood that the future aggregate load exceeds a consumption threshold during the operating cycle of the requested device. Based on the outcome, the control agent initiates or defers the activation request. Using three consumer reference models, a case study assessing EMS performance with respect to model detection, state estimation, and control as a function of consumer comfort and grid-informed consumption constraints is presented. A tradeoff analysis between comfort, consumption threshold, and appliance activation delay is demonstrated.
The Internet of Things envisions a web-connected infrastructure of sensing and actuation devices. However, the current state of the art presents another reality: monolithic end-to-end applications tightly coupled to a limited set of sensors and actuators. Growing such applications with new devices or behaviors, or extending the existing infrastructure with new applications, involves redesign and deployment. A proposed approach breaks these applications up into an equivalent set of functional units called context engines, whose I/O transformations are driven by general-purpose machine learning. This approach decreases computational redundancy and complexity with a minimal impact on accuracy. Researchers evaluated this approach's scalability--how the context engines' overhead grows as the input data and number of computational nodes increase. In a large-scale case study of residential smart-grid control, this approach provided better accuracy and scaling than the state-of-the-art single-stage approach.
We consider the problem of generating randomized control sequences for
complex networked systems typically actuated by human agents. Our approach
leverages a concept known as control improvisation, which is based on a
combination of data-driven learning and controller synthesis from formal
specifications. We learn from existing data a generative model (for instance,
an explicit-duration hidden Markov model, or EDHMM) and then supervise this
model in order to guarantee that the generated sequences satisfy some desirable
specifications given in Probabilistic Computation Tree Logic (PCTL). We present
an implementation of our approach and apply it to the problem of mimicking the
use of lighting appliances in a residential unit, with potential applications
to home security and resource management. We present experimental results
showing that our approach produces realistic control sequences, similar to
recorded data based on human actuation, while satisfying suitable formal
requirements.
The goal of the Smart* project is to optimize home energy con-sumption. As part of the project, we have designed and deployed a "live" system that continuously gathers a wide variety of envi-ronmental and operational data in three real homes. In contrast to prior work, our focus has been on sensing depth, i.e., collecting as much data as possible from each home, rather than breadth, i.e., collecting data from as many homes as possible. Our data captures many important aspects of the home environment, including aver-age household electricity usage every second, as well as usage at every circuit and nearly every plug load, electricity generation data from on-site solar panels and wind turbines, outdoor weather data, temperature and humidity data in indoor rooms, and, finally, data for a range of important binary events, e.g., at wall switches, the HVAC system, doors, and from motion sensors. We also have elec-tricity usage data every minute from 400 anonymous homes. This data corpus has served as the foundation for much of our recent research. In this paper, we describe our data sets as well as basic software tools we have developed to facilitate their collection. We are releasing both the data and tools publicly to the research com-munity to foster future research on designing sustainable homes.
Detailed information about a home's occupancy is necessary to implement many advanced energy-efficiency optimizations. However, monitoring occupancy directly is intrusive, typically requiring the deployment of multiple environmental sensors, e.g., motion, acoustic, CO2, etc. In this paper, we explore the potential for Non-Intrusive Occupancy Monitoring (NIOM) by using electricity data from smart meters to infer occupancy. We first observe that a home's pattern of electricity usage generally changes when occupants are present due to their interact with electrical loads. We empirically evaluate these interactions by monitoring ground truth occupancy in two homes, then correlating it with changes in statistical metrics of smart meter data, such as power's mean and variance, over short intervals. In particular, we use each metric's maximum value at night as a proxy for its maximum value in an unoccupied home, and then signal occupancy whenever the daytime value exceeds it. Our results highlight NIOM's potential and its challenges.
This paper presents a novel approach to derive U.S. residential building energy load profiles. This approach uses bootstrap sampling method to extract daily activity pattern of occupants of a household from American Time Use Survey (ATUS) data. The characteristics of ATUS data, the relation between time-use and load-demand, and the robustness of this approach are discussed. Virtual experiments were conducted on Energy Plus platform to study the patterns of annual load demand distribution under different household composition and thermal zoning schemes. Simulations of average 24-hr appliance and lighting load profiles were also conducted. The simulated load profiles and those from utility metering studies have good agreement. This novel approach has versatile applications in residential building energy simulation.
This paper presents a model to simulate the electricity demand of a single household consisting of multiple individuals. The total consumption is divided into four main categories, namely cold appliances, heating, ventilation, and air conditioning, lighting, and energy consumed by household members’ activities. The first three components are modeled using engineering physically-based models, while the activity patterns of individuals are modeled using a heterogeneous Markov chain. Using data collected by the U.S. Bureau of Labor Statistics, a case study for an average American household is developed. The data are used to conduct an in-sample validation of the modeled activities and a rigorous statistical validation of the predicted electricity demand against metered data is provided. The results show highly realistic patterns that capture annual and diurnal variations, load fluctuations, and diversity between household configuration, location, and size.
The Smart Grid is drawing attention from various research areas. Distributed control algorithms at different scales within the grid are being developed and deployed; yet their effects on each other and the grid's health and stability has not been sufficiently studied due to the lack of a capable simulator. Simulators in the literature can solve the power flow by modeling the physical system, but fail to address the cyber physical aspect of the smart grid with multiple agents. To answer these questions, we have developed S2Sim: Smart Grid Swarm Simulator. S2Sim allows any object within the grid to have its own independent control, transforming physical elements into cyber-physical representations. Objects can have any size ranging from a light bulb to a whole microgrid and their representative data can be supplied from a real device, simulation, distributed control algorithm or a database. S2Sim shields the complexity of the power flow solution from the control algorithms and directly supplies information on system stability. This information can be used to give feedback signals like price or regulation incentives by virtual coordinators to form closed-loop control. Using three case studies, we illustrate how different distributed control algorithms can have varying effects on system stability, which would go undetected in the absence of our simulator. Furthermore, our case studies show that control algorithms can't be justified without being tested within the grid.
The growing energy consumption in the residential sector represents about 30% of global demand. This calls for Demand Side Management solutions propelling change in behaviors of end consumers, with the aim to reduce overall consumption as well as shift it to periods in which demand is lower and where the cost of generating energy is lower. Demand Side Management solutions require detailed knowledge about the patterns of energy consumption. The profile of electricity demand in the residential sector is highly correlated with the time of active occupancy of the dwellings; therefore in this study the occupancy patterns in Spanish properties was determined using the 2009-2010 Time Use Survey (TUS), conducted by the National Statistical Institute of Spain. The survey identifies three peaks in active occupancy, which coincide with morning, noon and evening. This information has been used to input into a stochastic model which generates active occupancy profiles of dwellings, with the aim to simulate domestic electricity consumption. TUS data were also used to identify which appliance-related activities could be considered for Demand Side Management solutions during the three peaks of occupancy.
This paper presents a statistical method for modeling the behavior of household occupants to estimate residential energy consumption. Using data gathered by the U.S. Census Bureau in the American Time Use Survey (ATUS), actions carried out by survey respondents are categorized into ten distinct activities. These activities are defined to correspond to the major energy consuming loads commonly found within the residential sector. Next, time varying minute resolution Markov chain based statistical models of different occupant types are developed. Using these behavioral models, individual occupants are simulated to show how an occupant interacts with the major residential energy consuming loads throughout the day. From these simulations, the minimum number of occupants, and consequently the minimum number of multiple occupant households, needing to be simulated to produce a statistically accurate representation of aggregate residential behavior can be determined. Finally, future work will involve the use of these occupant models along side residential load models to produce a high-resolution energy consumption profile and estimate the potential for demand response from residential loads.
Energy and sustainability issues raise a large number of problems that can be tackled using approaches from data mining and machine learning, but traction of such problems has been slow due to the lack of publicly available data. In this paper we present the Reference Energy Disaggregation Data Set (REDD), a freely available data set containing de-tailed power usage information from several homes, which is aimed at furthering research on energy disaggregation (the task of determining the component appliance contributions from an aggregated electricity signal). We discuss past ap-proaches to disaggregation and how they have influenced our design choices in collecting data, we describe the hardware and software setups for the data collection, and we present initial benchmark disaggregation results using a well-known Factorial Hidden Markov Model (FHMM) technique.
Residential energy constitutes 38% of the total energy consumption in the United States [1]. Although a number of building simulators have been proposed, there are no residential electrical energy simulators capable of modeling complex scenarios and exploring the tradeoffs in home energy management. We propose HomeSim, a residential electrical energy simulation platform that enables investigating the impact of technologies such as renewable energy and different battery types. Additionally, HomeSim allows us to simulate different scenarios including centralized vs. distributed in-home energy storage, intelligent appliance rescheduling, and outage management. Using measured residential data, HomeSim quantifies different benefits for different technologies and scenarios, including up to 50% reduction in grid energy through a combination of distributed batteries and reschedulable appliances.
This paper presents a Markov chain approach for developing low-voltage (LV) residential load models for use in the analysis of future smart grids. The motivation is to obtain a better understanding of load use and user behaviour at the LV level which will allow future electricity network performance to be improved by understanding the impact of energy demand transformations through the application of demand-side management (DSM) strategies at the LV side. The developed load models highlight the variations between individual and aggregate load model representation and show that the composition of LV residential users should be carefully considered when analysing LV networks.
We present a novel energy management system for residential demand response. The algorithm, named CAES, reduces residential energy costs and smooths energy usage. CAES is an online learning application that implicitly estimates the impact of future energy prices and of consumer decisions on long term costs and schedules residential device usage. CAES models both energy prices and residential device usage as Markov, but does not assume knowledge of the structure or transition probabilities of these Markov chains. CAES learns continuously and adapts to individual consumer preferences and pricing modifications over time. In numerical simulations CAES reduced average end-user financial costs from 16% to 40% with respect to a price-unaware energy allocation.