Jack Kelly

NILMTK: an open source toolkit for non-intrusive load monitoring

Non-intrusive load monitoring, or energy disaggregation, aims to separate household energy consumption data collected from a single point of measurement into appliance-level consumption data. In recent years, the field has rapidly expanded due to increased interest as national deployments of smart meters have begun in many countries. However, empirically comparing disaggregation algorithms is currently virtually impossible. This is due to the different data sets used, the lack of reference implementations of these algorithms and the variety of accuracy metrics employed. To address this challenge, we present the Non-intrusive Load Monitoring Toolkit (NILMTK); an open source toolkit designed specifically to enable the comparison of energy disaggregation algorithms in a reproducible manner. This work is the first research to compare multiple disaggregation approaches across multiple publicly available data sets. Our toolkit includes parsers for a range of existing data sets, a collection of preprocessing algorithms, a set of statistics for describing data sets, two reference benchmark disaggregation algorithms and a suite of accuracy metrics. We demonstrate the range of reproducible analyses which are made possible by our toolkit, including the analysis of six publicly available data sets and the evaluation of both benchmark disaggregation algorithms across such data sets.

The UK-DALE dataset, domestic appliance-level electricity demand and whole-house demand from five UK homes

Many countries are rolling out smart electricity meters. These measure a home’s total power demand. However, research into consumer behaviour suggests that consumers are best able to improve their energy efficiency when provided with itemised, appliance-by-appliance consumption information. Energy disaggregation is a computational technique for estimating appliance-by-appliance energy consumption from a whole-house meter signal. To conduct research on disaggregation algorithms, researchers require data describing not just the aggregate demand per building but also the ‘ground truth’ demand of individual appliances. In this context, we present UK-DALE: an open-access dataset from the UK recording Domestic Appliance-Level Electricity at a sample rate of 16 kHz for the whole-house and at 1/6 Hz for individual appliances. This is the first open access UK dataset at this temporal resolution. We recorded from five houses, one of which was recorded for 655 days, the longest duration we are aware of for any energy dataset at this sample rate. We also describe the low-cost, open-source, wireless system we built for collecting our dataset.

Metadata for Energy Disaggregation

Energy disaggregation is the process of estimating the energy consumed by individual electrical appliances given only a time series of the whole-home power demand. Energy disaggregation researchers require datasets of the power demand from individual appliances and the whole-home power demand. Multiple such datasets have been released over the last few years but provide metadata in a disparate array of formats including CSV files and plain-text README files. At best, the lack of a standard metadata schema makes it unnecessarily time-consuming to write software to process multiple datasets and, at worse, the lack of a standard means that crucial information is simply absent from some datasets. We propose a metadata schema for representing appliances, meters, buildings, datasets, prior knowledge about appliances and appliance models. The schema is relational and provides a simple but powerful inheritance mechanism.

Dataport and NILMTK: A building data set designed for non-intrusive load monitoring

Non-intrusive load monitoring (NILM), or energy disaggregation, is the process of using signal processing and machine learning to separate the energy consumption of a building into individual appliances. In recent years, a number of data sets have been released in order to evaluate such approaches, which contain both building-level and appliance-level energy data. However, these data sets typically cover less than 10 households due to the financial cost of such deployments, and are not released in a format which allows the data sets to be easily used by energy disaggregation researchers. To this end, the Dataport database was created by Pecan Street Inc, which contains 1 minute circuit-level and building-level electricity data from 722 households. Furthermore, the non-intrusive load monitoring toolkit (NILMTK) was released in 2014, which provides software infrastructure to support energy disaggregation research, such as data set parsers, benchmark disaggregation algorithms and accuracy metrics. This paper describes the release of a subset of the Dataport database in NILMTK format, containing one month of electricity data from 669 households. Through the release of this Dataport data in NILMTK format, we pose a challenge to the signal processing community to produce energy disaggregation algorithms which are both accurate and scalable.

NILMTK v0.2: a non-intrusive load monitoring toolkit for large scale data sets: demo abstract

In this demonstration, we present an open source toolkit for evaluating non-intrusive load monitoring research; a field which aims to disaggregate a households total electricity consumption into individual appliances. The toolkit contains: a number of importers for existing public data sets, a set of preprocessing and statistics functions, a benchmark disaggregation algorithm and a set of metrics to evaluate the performance of such algorithms. Specifically, this release of the toolkit has been designed to enable the use of large data sets by only loading individual chunks of the whole data set into memory at once for processing, before combining the results of each chunk.

Characterization of phosphodiesterase 4 in guinea‐pig macrophages: multiple activities, association states and sensitivity to selective inhibitors

The cyclic AMP phosphodiesterases (PDE) in guinea‐pig peritoneal macrophages were isolated, partially characterized and their role in regulating the cyclic AMP content in intact cells evaluated. Differential centrifugation of macrophage lysates revealed that ∼90% of the PDE activity was membrane‐bound and exclusively hydrolyzed cyclic AMP. This activity was not removed by KCl (200 mM) but was readily solubilized by the non‐ionic detergent, Triton X‐100 (1% v/v). Greater than 80% of the hydrolytic activity was suppressed by the PDE4 inhibitors, R‐rolipram and nitraquazone with IC50s of 240 and 540 nM, respectively. Anion‐exchange chromatography of the total protein extracted from macrophages resolved two major peaks of cyclic AMP PDE activity that were insensitive to cyclic GMP (10 μM), calmodulin (50 units plus 2 mM CaCl2) and a PDE3 inhibitor, SK&F 95654 (10 μM), but were markedly suppressed by RS‐rolipram (10 μM). The two peaks of PDE activity were arbitrarily designated CPPDE4α and CPPDE4β with respect to the order from which they were eluted from the column where the prefix, CP, refers to the species, Cavia porcellus. The hydrolysis of cyclic AMP catalyzed by CPPDE4α and CPPDE4β conformed to Michaelis‐Menten kinetic behaviour with similar Kms (13.4 and 6.4 μM, respectively). Thermal denaturation of membrane‐bound PDE4 at 50°C followed bi‐exponential kinetics with t1/2 values of 1.5 and 54.7 min for the first and second components, respectively. In contrast, CPPDE4α and CPPDE4β each decayed mono‐exponentially with significantly different thermostabilities (t1/2=2.77 and 1.15 min, respectively). Gel filtration of CPPDE4β separated two peaks of rolipram‐sensitive PDE activity. The main peak eluted at a volume indicative of a ∼180 kDa protein but was preceded by a much larger form of the enzyme that had an estimated weight of 750 kDa. Size exclusion chromatography of CPPDE4α resolved a broad peak of activity with molecular weights spanning 50 to 200 kDa. Of ten PDE inhibitors examined, none distinguished CPPDE4α from CPPDE4β with respect to their IC50 values or their rank order of potency. RS‐rolipram acted as a purely competitive inhibitor of cyclic AMP hydrolysis with Kis of 2 μM and 1.5 μM for CPPDE4α and CPPDE4β, respectively. In contrast to the membrane‐associated enzyme(s), R‐rolipram and nitraquazone were 4 to 19 fold less potent as inhibitors of CPPDE4α and CPPDE4β. In intact macrophages, Ro 20‐1724 and RS‐rolipram potentiated isoprenaline‐induced cyclic AMP accumulation under conditions where a PDE3 inhibitor, SK&F 94120, was essentially inactive. These data demonstrate that the predominant cyclic AMP hydrolyzing activity in guinea‐pig macrophages is a PDE4. Moreover, thermostability studies and size exclusion chromatography indicates the possible expression of two intrinsic, membrane‐associated isoenzymes which can regulate the cyclic AMP content in intact cells. The finding that soluble and particulate forms of the same enzyme exhibit different sensitivities to rolipram and nitraquazone implies that PDE4 can change conformation. Finally, the identification of multiple molecular weight species of CPPDE4 suggests that this enzyme(s) might form multimeric complexes of variable association states.

Smarter Electricity and Argumentation Theory

The current, widespread introduction of smart electricity meters is resulting in large datasets’ becoming available, but there is as yet little in the way of advanced data analytics and visualization tools, or recommendation software for changes in contracts or user behaviour, which use this data. In this paper we present an integrated tool which combines the use of abstract argumentation theory with linear optimization algorithms, to achieve some of these ends.