Autonomous load disaggregation approach based on active power measurements

Abstract

With the help of smart metering valuable information of the appliance usage can be retrieved. In detail, non-intrusive load monitoring (NILM), also called load disaggregation, tries to identify appliances in the power draw of an household. In this paper an unsupervised load disaggregation approach is proposed that works without a priori knowledge about appliances. The proposed algorithm works autonomously in real time. The number of used appliances and the corresponding appliance models are learned in operation and are progressively updated. The proposed algorithm is considering each useful and suitable detected power state. The algorithm tries to detect power states corresponding to on/off appliances as well as to multi-state appliances based on active power measurements in 1s resolution. We evaluated the novel introduced load disaggregation approach on real world data by testing the possibility to disaggregate energy demand on appliance level.

On a Training-Less Solution for Non-Intrusive Appliance Load Monitoring Using Graph Signal Processing

“…The REDD dataset is widely used for the evaluation of various NALM approaches [10], [14], [16], [23], [31], [38]. REDD houses contain several appliances with a small unknown load.…”

“…In [37], an unsupervised low-rate NALM approach, based on clustering and matching pursuit is proposed; however, the approach uses both active and reactive power, performs poorly for appliance loads below 400W and concludes that the results might improve only if additional features are included. An unsupervised FHMM-based approach introduced in [38] learns the appliance models on-the-fly, thus its performance gradually improves requiring some time to reach high accuracy.…”

“…Since our approach is signal processing based (as opposed to machine learning based), unlike the method presented in [38], our approach does not need to learn appliance features over time, since it is not based on the appliance model generation.…”

Occupancy based household energy disaggregation using ultra wideband radar and electrical signature profiles

“…Egarter [22] describes a real time unsupervised load disaggregation algorithm which uses only real power values, sampled at 1Hz. The pattern recognition method used is particle filtering, which is computationally expensive.…”

Load Disaggregation Based on Aided Linear Integer Programming

“…The ultimate goal for NILM is to enable disaggregation without the need for supervised learning. There has been some recent progress in this area [10], [11], although the accuracy of such methods is still low. Alternative approaches such as combinatorial optimization or integer programming (IP) have been much less explored.…”