IEC TS 63297-2021 pdf free download – Sensing devices for non-intrusive load monitoring (NILM) systems.
4.2NILM sensing device
A NILM sensing device (NSD) is a device connected to the electrical installation. It producesdata that can be used by NILM analytics. Examples of data that may be produced by an NSD
include:
samples of current and/or voltage waveforms;
features characterizing the current and/or voltage waveforms;
features related to the high-frequency patterns in the electrical signals;estimated values of electrical parameters;
measured values of electrical parameters.4.3NILM analytics
The value of NILM systems is essentially in the
analytics and how well they are able to make
use of the data produced by the NSD.
NILM analytics are algorithms that analyse the data output by an NSD and produce estimateddisaggregated information that can help stakeholders make decisions.
Examples of information that may be produced by NILM analytics are:
estimated disaggregation of energy consumption into specific usages (heating, refrigeration,entertainment.. .);
estimated disaggregation of energy consumption into specific types of appliances (ovens,fridges, pumps…).
NOTE NILM systems produce estimated disaggregation into energy usages. When accurate measurement andanalysis of energy consumption andlor otherelectnical parameters is needed (e.g. for monitoring the electricalinstallation),systems based on standardized measuring devices (e.g.PMDs or meters) are used.
5 Classification of NILM sensing devices (NSD)
5.1 General
ANILM sensing device (NSD) is a gateway between the physical electrical installation and theworld of analytics. ln order to operate efficiently, NlILM algorithms need to know the type of datathey are to process. The behaviour of the NSD depends on several characteristics (see Figure3).
5.2 Definition of NSD classes
NILM sensing devices should be classified according to a limited number of essential parameters, for example:
– Input sampling frequency: the frequency at which the electrical signals are sampled by the NSD. This frequency typically varies from a couple of kHz to the MHz range. The sampling process may be gapless or not.
This parameter is essential to characterize NSDs that produce samples of the electrical waveforms. It is less relevant for NSDs producing estimated or measured values; therefore a separate class is specified (class P as “parameters only”).
– Output rate: the rate at which the NSD produces data that can be used by NILM analytics.
This rate typically varies from one set of data per second to one set of data per 30 minutes.
– Data bit rate: the average bit-per-second (bps) over an hour at which the electrical signals are quantified by the NSD. This data bit rate typically varies from a few bps to the Mbps range.
Table 1, Table 2 and Table 3 provide a classification of NILM sensing devices according to these three parameters.
Class P in Table 1 is for NSDs that do not produce samples of the electrical waveforms. Instead,they produce measurements or estimates of electrical parameters (e.g. active power, reactive power, power factor, harmonic distortion, etc.).
– Classical measuring instruments like power metering and monitoring devices (PMD) can be considered class-P NSDs as they produce measurements of electrical parameters.
– Power quality instruments (PQI) also produce measurements of electrical parameters, but also often have waveform capture capabilities. They can therefore be considered as class-P NSDs and, typically, class 3 or 4 NSDs depending on their sampling properties.
The classes from Table 1, Table 2 and Table 3 should be combined. For example,IEC TS 63297 pdf download.