1. Data Preprocessing

How the data generated by the Smart Meter is reprocessed for learning

Notations and Definitions

1. Raw reading Data

A Raw Data Objec

• t : Timestamp

• P(t) : Power at that time

• S(t) : Apparent power at that time

• V(t) : Mains root mean squared (RMS) at that time

2. Time Series

Time series is a sequence of Raw Data Objects R. The length of a Time series is defiend by the Time value of the last Data Object in T minus the first on. So Tn- T1.

A time Series Object

3. Subtime Series

A Subtime Series is defined for a Time Series, as a continuous subsequence of T. You can imagine it , as just a slice of a Timeseries(?).

A Subtimeseries Object

4. Sliding time window

The sliding window represents a set of subtime series of T. It is the fundamental timing component of the learning engine

A Sliding window Object