A STEADY-STATE POWER GENERATION METHOD AND STACKED LSTM IN EVENT DETECTION FOR HOUSEHOLD APPLIANCES
Abstract
Non-Intrusive Load Monitoring (NILM) is an energy monitoring system that has gained popularities in recent years for improving energy saving. NILM for home electricity management provides useful information on home appliances, for example, estimating electricity consumption of individual appliances from electric power measurement and habits of consumers in using electricity. Moreover, this system has low-cost metering solution that makes research community and companies more interest to develop it. In order to reduce sensing infrastructure costs, NILM monitors the electric load based on machine learning methods using only one sensor device. Besides, the event detection method is one of the cores in NILM that can accurately determine which appliance is ON or OFF within a period of time. The paper presents a new event detection method in a low sampling rate. This proposed method includes three main parts: (1) Generating the steady-state power signal based on the CUSUM (Cumulative SUM) signal extracted from the active power signal; (2) Generating mean and variance signals based on the generated steady-state power signal; (3) Applying Stacked LSTM (Stacked Long Short-Term Memory) model to improve event detection performance on the extracted features. The experiments are performed on two public datasets which are AMPds2 (The Almanac of Minutely Power dataset (Version 2)), UK-DALE (UK Domestic Appliance-Level Electricity). The experimental results indicate that the proposed method achieves high performance accuracy for appliance event detection by 94% to 100% in terms of the Receiver Operating Characteristic (ROC) curve.
References
Hart, G.W. Nonintrusive appliance load monitoring. IEEE Proc., 1992, 80, pp.1870–1891.
Zoha, A.; Gluhak, A.; Imran, M.A.; Rajasegarar, S. Non-Intrusive Load Monitoring Approaches for Disaggregated Energy Sensing: A Survey. Sensors, 2012, 12, pp. 16838–16866.
Klemenjak, C., Goldsborough, P. Non-Intrusive Load Monitoring: A Review and Outlook. ArXiv, https://arxiv.org/pdf/1610.01191.pdf, 2016, pp. 1–17.
Faustine, A., et al. A Survey on Non-Intrusive Load Monitoring Methodies and Techniques for Energy Disaggregation Problem. ArXiv, https://arxiv.org/pdf/1703.00785.pdf, 2017, pp.1–17.
Norford, L.K.; Leeb, S.B. Non-intrusive electrical load monitoring in commercial buildings based on steady-state and transient load-detection algorithms. Energ. Build, 1996, 24, pp.51–64.
Marceau, M.L.; Zmeureanu, R. Nonintrusive load disaggregation computer program to estimate the energy consumption of major end uses in residential buildings. Energ. Convers. Manag, 2000, 41, pp.1389–1403.
Marchiori, A.; Hakkarinen, D.; Han, Q.; Earle, L. Circuit-level load monitoring for household energy management. IEEE Pervas. Comput., 2011, 10, pp.40–48.
He X., Pun M., and Kou C.C.J. Change-point detection-based power quality monitoring in smart grids. SIP 2015, vol. 4, pp. 1–9.
Zhu, Z., Zhang S., Wei, Z., Yin, B., Huang, X. A novel CUSUM-based approach for event detection in smart metering. OP Conf. Ser.: Mater. Sci. Eng., 2018; doi:10.1088/1757-899X/322/7/072014.
T. Le, T. Park, D. Cho and H. Kim, “An Effective Classification for DoS Attacks in Wireless Sensor Networks,” 2018 Tenth International Conference on Ubiquitous and Future Networks (ICUFN), Prague, 2018, pp. 689-692; doi: 10.1109/ICUFN.2018.8436999.
Nguyen Duy Tan, Nguyen Thi Thanh Hue, and Vu Thi Thu Huong, Simulation and Evaluation about MAC Protocols in Wireless Sensor Network Using NS2. UTEHY Journal of Science and Technology, vol 9, no. 23, pp. 39-45.
Kelly, J., Knotttenbelt, W. Neural NILM: Deep Neural Networks Applied to Energy Disaggregation. arXiv.org, 2015, https://arxiv.org/pdf/1507.06594.pdf.
Kelly, J.D.: Disaggregation of domestic smart meter energy data. Ph.D thesis, August 2016.
Paulo, P.: Applications of deep learning techniques on NILM. Ph.D thesis, 2016.
J., Kim, T.T.H, Le, and H., Kim, Nonintrusive Load Monitoring Based on Advanced Deep Learning and Novel Signature, Computational Intelligence and Neuroscience 2017, Article ID 4216281, 22 pages; https://doi.org/10.1155/2017/4216281.
Le, T.-T.-H.; Kim, H. Non-Intrusive Load Monitoring Based on Novel Transient Signal in Household Appliances with Low Sampling Rate. Energies, 2018, 11, 3409.
Rafig, H., et al. Regularized LSTM Based Deep Learning Model: Fist Step towards Real-Time Non-Intrusive Load Monitoring. 2018 the 6th International Conference on Smart Energy Grid Engineering.
T.-T.-H. Le, J. Kim and H. Kim. Classification performance using gated recurrent unit recurrent neural network on energy disaggregation. 2016 International Conference on Machine Learning and Cybernetics (ICMLC), Jeju, 2016, pp. 105-110; doi: 10.1109/ICMLC.2016.7860885.
Liu, B.; Fu, C.; Bielefield, A.; Liu, Y.Q. Forecasting of Chinese Primary Energy Consumption in 2021 with GRU Artificial Neural Network. Energies 2017, 10, 1453.
T. Le, H. Kang and H. Kim, “Household Appliance Classification Using Lower Odd-Numbered Harmonics and the Bagging Decision Tree,” in IEEE Access, 2020, vol. 8, pp. 55937-55952; doi: 10.1109/ACCESS.2020.2981969.
Mukaroh, A.; T.T.H., Le, and Kim, H. Background Load Denoising across Complex Load Based on Generative Adversarial Network to Enhance Load Identification. Sensors, 2020, 20, 5674.
S. Makonin, B. Ellert, I. V. Bajic, and F. Popowich. Electricity, water, and natural gas consumption of a residential house in Canada from 2012 to 2014. Scientific Data 2016, vol. 3, no. 160037, pp. 1–12
Kelly L., Knottenbelt, W. The UK-DALE dataset, domestic appliance-level electricity demand and whole-house demand from five UK homes. Scientific Data 2 2015, Article number:150007; DOI:10.1038/sdata.2015.7.
Makonin, S., et al. AMPds: A Public Dataset for Load Disaggregation and Eco-Feedback Research. EPEC, 2013, pp.1–6.
