An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring
Non-Intrusive Load Monitoring (NILM) or load disaggregation aims to analyze power consumption by decomposing the energy measured at the aggregate level into constituent appliances level. The conventional load disaggregation framework consists of signal processing and machine learning-based pipelined...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
D. G. Pylarinos
2021-06-01
|
| Series: | Engineering, Technology & Applied Science Research |
| Subjects: | |
| Online Access: | https://www.etasr.com/index.php/ETASR/article/view/4142 |
| _version_ | 1798017389442367488 |
|---|---|
| author | S. Wali M. H. U. Haq M. Kazmi S. A. Qazi |
| author_facet | S. Wali M. H. U. Haq M. Kazmi S. A. Qazi |
| author_sort | S. Wali |
| collection | DOAJ |
| description | Non-Intrusive Load Monitoring (NILM) or load disaggregation aims to analyze power consumption by decomposing the energy measured at the aggregate level into constituent appliances level. The conventional load disaggregation framework consists of signal processing and machine learning-based pipelined architectures, respectively for explicit feature extraction and decision making. Manual feature selection in such load disaggregation frameworks leads to biased decisions that eventually reduce system performance. This paper presents an efficient End-to-End (E2E) approach-based unified architecture using Gated Recurrent Units (GRU) for NILM. The proposed approach eliminates explicit feature engineering and has a unified classification and prediction model for appliance power. This eventually reduces the computational cost and enhances response time. The performance of the proposed system is compared with conventional algorithms' with the use of recall, precision, accuracy, F1 score, the relative error in total energy and Mean Absolute Error (MAE). These evaluation metrics are calculated on the power consumption of top priority appliances of Reference Energy Disaggregation Dataset (REDD). The proposed architecture with an overall accuracy of 91.2 and MAE of 25.23 outperforms conventional methods for all electrical appliances. It has been showcased through a series of experiments that feature extraction and event-based approaches for NILM can readily be replaced with E2E deep learning techniques allowing simpler and cost-efficient implementation pathways. |
| first_indexed | 2024-04-11T16:06:28Z |
| format | Article |
| id | doaj.art-aa57e94d08f7461da10fca0930b52a5a |
| institution | Directory Open Access Journal |
| issn | 2241-4487 1792-8036 |
| language | English |
| last_indexed | 2024-04-11T16:06:28Z |
| publishDate | 2021-06-01 |
| publisher | D. G. Pylarinos |
| record_format | Article |
| series | Engineering, Technology & Applied Science Research |
| spelling | doaj.art-aa57e94d08f7461da10fca0930b52a5a2022-12-22T04:14:49ZengD. G. PylarinosEngineering, Technology & Applied Science Research2241-44871792-80362021-06-0111310.48084/etasr.4142An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load MonitoringS. Wali0M. H. U. Haq1M. Kazmi2S. A. Qazi3Faculty of Electrical and Computer Engineering, NED University of Engineering & Technology, PakistanFaculty of Electrical and Computer Engineering, NED University of Engineering & Technology, PakistanFaculty of Electrical and Computer Engineering, NED University of Engineering & Technology, PakistanNeurocomputation Lab, National Centre of Artificial Intelligence, NED University of Engineering and Technology, PakistanNon-Intrusive Load Monitoring (NILM) or load disaggregation aims to analyze power consumption by decomposing the energy measured at the aggregate level into constituent appliances level. The conventional load disaggregation framework consists of signal processing and machine learning-based pipelined architectures, respectively for explicit feature extraction and decision making. Manual feature selection in such load disaggregation frameworks leads to biased decisions that eventually reduce system performance. This paper presents an efficient End-to-End (E2E) approach-based unified architecture using Gated Recurrent Units (GRU) for NILM. The proposed approach eliminates explicit feature engineering and has a unified classification and prediction model for appliance power. This eventually reduces the computational cost and enhances response time. The performance of the proposed system is compared with conventional algorithms' with the use of recall, precision, accuracy, F1 score, the relative error in total energy and Mean Absolute Error (MAE). These evaluation metrics are calculated on the power consumption of top priority appliances of Reference Energy Disaggregation Dataset (REDD). The proposed architecture with an overall accuracy of 91.2 and MAE of 25.23 outperforms conventional methods for all electrical appliances. It has been showcased through a series of experiments that feature extraction and event-based approaches for NILM can readily be replaced with E2E deep learning techniques allowing simpler and cost-efficient implementation pathways.https://www.etasr.com/index.php/ETASR/article/view/4142non-intrusive load monitoringgated recurrent unitsend-to-end machine learningreference energy disaggregation dataset |
| spellingShingle | S. Wali M. H. U. Haq M. Kazmi S. A. Qazi An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring Engineering, Technology & Applied Science Research non-intrusive load monitoring gated recurrent units end-to-end machine learning reference energy disaggregation dataset |
| title | An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring |
| title_full | An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring |
| title_fullStr | An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring |
| title_full_unstemmed | An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring |
| title_short | An End-to-End Machine Learning based Unified Architecture for Non-Intrusive Load Monitoring |
| title_sort | end to end machine learning based unified architecture for non intrusive load monitoring |
| topic | non-intrusive load monitoring gated recurrent units end-to-end machine learning reference energy disaggregation dataset |
| url | https://www.etasr.com/index.php/ETASR/article/view/4142 |
| work_keys_str_mv | AT swali anendtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT mhuhaq anendtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT mkazmi anendtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT saqazi anendtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT swali endtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT mhuhaq endtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT mkazmi endtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring AT saqazi endtoendmachinelearningbasedunifiedarchitecturefornonintrusiveloadmonitoring |