Data for: Multivariate Event Detection Methods for Non Intrusive Load Monitoring in Residential Buildings

The Power_time_series.txt file contains the 34 electrical features related to power that were computed using current and voltage acquisitions made using our own acquisition system based on an Arduino MKR Zero microcontroller with a sampling frequency of 6.25 kHz (see reference "Design of an electricity consumption measurement system for Non Intrusive Load Monitoring", IEEE IREC 2019 from the same authors).

In the text file the power time series correspond to the active power P, its harmonic order Pk, where k is ranging between 1 and 15 and the sum of the harmonics PH. There are also the reactive power Q, its harmonic order Qk, where k is ranging between 1 and 15 and the sum of the harmonics QH.

All these 34 features are ordered by columns as follows:
P, P1, P2, P3, P4, P5, P6, P7, P8, P9, P10, P11, P12, P13, P14, P15, PH, Q, Q1, Q2, Q3, Q4, Q5, Q6, Q7, Q8, Q9, Q10, Q11, Q12, Q13, Q14, Q15, QH

These power time series correspond to the same scenario of 12 appliances (microwave, a DVD player, a fan, a screen, a vacuum, a waffle iron, a hair dryer, an iron, a flat iron, a mixer, a CFL and a LED lamp) that are randomly switched on/off every 3 seconds during almost one hour and half. This results in almots 1200 events of diferrent appliances.

The text file Events_time_stamping.txt contains the time stamps of each events that were hand-labelled after inspecting the signals. Timestamps of the labeled events were adjusted to match the transitions.

The text file Events_labelling corresponds to the labelling of each events (which appliance is responsable of the event).

Finally, the file MultivriateAbruptChangeDetectors.pdf is a repport containing the mathematicla derivation of each investigated detector in details.