A New Approach for Supervised Power Disaggregation by Using a Denoising Autoencoder and Recurrent LSTM Network

Wang, T. S.; Ji, Tianyao; Li, M. S.

doi:10.1109/demped.2019.8864870

Cited by 20 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wang et al [67] proposed an end-to-end method to identify individual appliances from aggregated data using a combination of DAE and LSTM networks on an AMPds dataset. The method was trained on aggregated data and tested on synthesised data.…”

Section: Deep Supervised Learningmentioning

confidence: 99%

Advances in Machine-Learning Based Disaggregation of Building Heating Loads: A Review

Lien,

Najafi,

Rajasekharan

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

This review article investigates the methods proposed for disaggregating the space heating units' load from the aggregate electricity load of commercial and residential buildings. It explores conventional approaches together with those that employ traditional machine learning, deep supervised learning and reinforcement learning. The review also outlines corresponding data requirements and examines the suitability of a commonly utilised toolkit for disaggregating heating loads from lowfrequency aggregate power measurements. It is shown that most of the proposed approaches have been applied to high-resolution measurements and that few studies have been dedicated to low-resolution aggregate loads (e.g. provided by smart meters). Furthermore, only a few methods have taken account of special considerations for heating technologies, given the corresponding governing physical phenomena. Accordingly, the recommendations for future works include adding a rigorous preprocessing step, in which features inspired by the building physics (e.g. lagged values for the ambient conditions and values that represent the correlation between heating consumption and outdoor temperature) are added to the available input feature pool. Such a pipeline may benefit from deep supervised learning or reinforcement learning methods, as these methods are shown to offer higher performance compared to traditional machine learning algorithms for load disaggregation.

show abstract

Section: Deep Supervised Learningmentioning

confidence: 99%

Advances in Machine-Learning Based Disaggregation of Building Heating Loads: A Review

Lien,

Najafi,

Rajasekharan

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…In the work by Bonfigli et al (2018) load disaggregation problem was treated as a noise reduction problem and they presented a DNN architecture based on DAE. A framework to identify individual appliances from the aggregated data based on DAE and LSTM was presented by Wang et al (2019). The DAE was utilized to reconstruct the single appliance signal and LSTM was applied to identify that the signal belongs to which electrical appliance.…”

Section: Overview Of Nilm Techniquesmentioning

confidence: 99%

Disaggregation of digital meter data for synthetic load profile generation

Bogaerts¹,

Jacobs²,

Ghane³

et al. 2021

Building Simulation Conference Proceedings

View full text Add to dashboard Cite

The electrical consumption has to be taken into account in building simulations. Empirically-based profiles are required, which can be generated by central measurements and using non-intrusive load monitoring (NILM) for disaggregation. In this work, we present an overview of NILM techniques, a comparison between two frequently used deep neural networks for individual appliance identification and we investigate the influence of the sampling rate with regards to the accuracy. Our best performing neural network is a combination of convolution and long-short-term memory networks. Furthermore, the sampling rate has a significant influence on the performance of neural networks in this context. There should be a trade-off between sampling rate and efficiency when applied in real-world devices. Key Innovations• A comparison between frequently used deep neural networks for individual appliance identification • An overview of Non-Intrusive Load monitoring techniques • Investigation into the relationship between sampling rate and accuracy for individual appliance identification when using deep learning starting from the current implemented 0,1Hz up to 60Hz. Practical ImplicationsThe combination of convolution neural networks and long-short term memory neural networks outperforms the state-of-the-art non-intrusive load monitoring (NILM) techniques for disaggregating the central measured electricity consumption.

show abstract

“…The performance of the combinatorial optimization (CO) based load disaggregation approach is compared with the approaches that were trained on the same dataset and reported in the literature [3] and [51]. Table 7.5 and figure 7.11 illustrate this comparison in terms of F1 score and it can be seen that the used CO has shown better disaggregation for two-state appliances such as heat pump but provided less disaggregate score for multistate appliances such as the fridge.…”

Section: Comparison Of Resultsmentioning

confidence: 99%

“…(2) The training set is made with the synthetic data (total power) which is in the function of the energy consumption of 7 important appliances that are present in a single house of AMPds. The entire training set can be defined as[51]:= ( 1 , 2 , 2 , 3 , . .…”

mentioning

confidence: 99%

Using Machine Learning and Deep Learning for Load disaggregation and Recognition of Activities in Household

Bimenyimana¹

View full text Add to dashboard Cite

A New Approach for Supervised Power Disaggregation by Using a Denoising Autoencoder and Recurrent LSTM Network

Cited by 20 publications

References 7 publications

Advances in Machine-Learning Based Disaggregation of Building Heating Loads: A Review

Advances in Machine-Learning Based Disaggregation of Building Heating Loads: A Review

Disaggregation of digital meter data for synthetic load profile generation

Using Machine Learning and Deep Learning for Load disaggregation and Recognition of Activities in Household

Contact Info

Product

Resources

About