Training load monitoring algorithms on highly sub-metered home electricity consumption data

Bergés, Mario; Goldman, Ethan; Matthews, H. Scott; Soibelman, Lúcio

doi:10.1016/s1007-0214(08)70182-2

Cited by 45 publications

(29 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While there have been numerous approaches to NILM [40,12,33,6,7,15,11,34,8,26], the most successful in residential settings are based on edge-detection [16,25]. Edge-detection techniques look for significant changes in the steady state current being consumed by the household.…”

Section: Introductionmentioning

confidence: 99%

Minimizing private data disclosures in the smart grid

Yang

et al. 2012

Proceedings of the 2012 ACM Conference on Computer and Communications Security

116

100

View full text Add to dashboard Cite

Smart electric meters pose a substantial threat to the privacy of individuals in their own homes. Combined with non-intrusive load monitors, smart meter data can reveal precise home appliance usage information. An emerging solution to behavior leakage in smart meter measurement data is the use of battery-based load hiding. In this approach, a battery is used to store and supply power to home devices at strategic times to hide appliance loads from smart meters. A few such battery control algorithms have already been studied in the literature, but none have been evaluated from an adversarial point of view. In this paper, we first consider two well known battery privacy algorithms, Best Effort (BE) and Non-Intrusive Load Leveling (NILL), and demonstrate attacks that recover precise load change information, which can be used to recover appliance behavior information, under both algorithms. We then introduce a stepping approach to battery privacy algorithms that fundamentally differs from previous approaches by maximizing the error between the load demanded by a home and the external load seen by a smart meter. By design, precise load change recovery attacks are impossible. We also propose mutual-information based measurements to evaluate the privacy of different algorithms. We implement and evaluate four novel algorithms using the stepping approach, and show that under the mutual-information metrics they outperform BE and NILL.

show abstract

Section: Introductionmentioning

confidence: 99%

Minimizing private data disclosures in the smart grid

Yang

et al. 2012

Proceedings of the 2012 ACM Conference on Computer and Communications Security

116

100

View full text Add to dashboard Cite

show abstract

“…Azar and Menassa (2011) have studied the residents' impact on energy use in commercial buildings, while Berges et al (2008Berges et al ( , 2011 were focused on load monitoring of home's appliances mainly in residential buildings. Mardookhy et al (2014) studied energy efficiency in residential buildings in Knoxville, while Agüero-Rubio et al (2014) proposed a methodology for management of thermal loads with real-time prices.…”

Section: Introductionmentioning

confidence: 99%

Wind energy integration via residential appliances

Xydis

Mihet‐Popa

2016

Energy Efficiency

View full text Add to dashboard Cite

In this paper cold and thermal storage options via wind energy surplus are examined aiming to enable higher wind integration. A case study in Greece with an installed wind farm (WF) was investigated reviewing the possibilities for higher renewables integration using thermostatically controlled loads (TCLs), analysing thermal and cold storage possibilities according to the needs of the local municipalities. The basic aim is to describe the idea of an aggregated system with a number of residential freezers, in order the wind power to be stored in the residential cold stores aiming at maximizing the benefit to the grid, utility or cold store owner avoiding at the same time wind curtailment. In addition, for thermal storage, by calculating thermal losses of the building, ventilation losses and adding up internal thermal gains of the building the space heating needs can be calculated.

show abstract

“…The proposed work intends to take energy feedback a step further by using RFI for appliance activity recognition and associating it to an end consumer. Some prior studies in appliance activity recognition have utilized distributed sensing approach by placing individual plug load monitors at all appliances to monitor individual power consumption [3]. Another technique proposed by Hart [4], utilizes disaggregated power information from NIALM based single point sensing method to do this.…”

Section: Introductionmentioning

confidence: 99%

Appliance Activity Recognition Using Radio Frequency Interference Emissions

2016

View full text Add to dashboard Cite

Over the past few decades, with rapid growth in infrastructure, there has been tremendous growth in energy consumption. Along with this, more and more electronic appliances are added to the existing infrastructure every day. Furthermore, the existing energy bills just provide an aggregate number of units consumed but fail to provide any actionable details of appliance level usage. With the quest for long-term energy sustainability and to reduce this ever-growing energy consumption, research groups across the globe have started looking into energy disaggregation as a means of providing feedback. Some promising techniques like Non-intrusive appliance load monitoring (NIALM) have been adopted to provide detailed energy breakdown to the end consumer. Despite all these efforts, energy attribution to the electrical activities still seems to be a farfetched goal, especially in shared spaces. In this work, we have analyzed the possibility of using RF emissions from electronic appliances to detect electrical activity. Besides their known operation, these appliances are known to radiate high-frequency noise in the ambient environment, also called radio frequency interference (or RFI). Hence by utilizing these RFI emissions from electronic appliances, electrical activity from the appliance can be detected in multiple frequency bands and at varying distances. An 8-fit Gaussian mixture model (GMM) and k-peak finder are used for feature extraction from RFI data, followed by appliance activity recognition using k-nearest neighbor based classification. Appliance detection is performed with a mean accuracy of 71.9% across 7-class classification problem. Finally, characteristic features of RFI observed from these appliances, are discussed.Index Terms-RFI (Radio frequency interference), kNN (knearest neighbor), NIALM (Non-intrusive appliance load monitoring), SMPS (Switched mode power supply), HF (High frequency), LF (Low frequency), UPS (Uninterruptable power supply), CFL (Compact fluorescent lamp). A 1530-437X (c)

show abstract

Training load monitoring algorithms on highly sub-metered home electricity consumption data

Cited by 45 publications

References 9 publications

Minimizing private data disclosures in the smart grid

Minimizing private data disclosures in the smart grid

Wind energy integration via residential appliances

Appliance Activity Recognition Using Radio Frequency Interference Emissions

Contact Info

Product

Resources

About