Creating a Baseline for Robust Energy Savings Estimation in Households

TL;DR

This paper introduces a methodology for establishing a user-specific hourly energy consumption baseline in households, enabling accurate energy savings estimation with minimal equipment, while addressing data quality issues and unordinary patterns.

Contribution

The paper presents a novel analytical approach for constructing year-long hourly energy baselines using low-cost sensors, accounting for data corruption and atypical consumption patterns.

Findings

Effective handling of missing and corrupted data.

Robust baseline construction for energy savings estimation.

Applicability to user motivation strategies.

Abstract

In this paper we present a methodology appropriate for establishing a user-specific hourly-based benchmark period of energy consumption. Such values can be used as reference for explicitly calculating energy savings. The required equipment is limited to low cost sensors. The so-called baseline allows for an explicit comparison of the household consumption now and then, once appropriate adjustments are made to handle the different conditions (e.g. temperature variation). When involving user motivation strategies, hourly based energy savings can provide multiple advantages. However, there are two major factors affecting the baseline construction: corrupted or missing data values and unordinary patterns. In this paper we provide an analytical methodology to handle such scenarios and create year-long hourly-based consumption baseline.