Non-intrusive load monitoring: Comparative analysis of transient state clustering methods

被引:6
作者
Etezadifar, Mozaffar [1 ]
Karimi, Houshang [1 ]
Mahseredjian, Jean [1 ]
机构
[1] Dept Elect Engn, Polytech Montreal, Montreal, PQ H3T 1J4, Canada
基金
加拿大自然科学与工程研究理事会;
关键词
Non-intrusive load monitoring; Unsupervised learning; Clustering; Transient state; Load disaggregation; Machine learning; NILM; ENERGY MANAGEMENT; EVENT DETECTION;
D O I
10.1016/j.epsr.2023.109644
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Non-intrusive load monitoring is one of the key tools in demand-side management (DSM). Recent advancements in the computational power of processors have accentuated the role of machine learning algorithms e.g., clustering, as a key function in the NILM solutions applied on power grids. In event-based NILM methods, the algorithm detects the transient states (load events) and clusters them based on the similarity of different features of the transient state. In this study, the performances of eight clustering algorithms are comprehensively investigated and the impact of choosing different input signals, e.g., P, Q, and I, on transient states clustering is analyzed. Various input signals from the BLUED dataset are fed to the clustering algorithms. By comparing the evaluation metrics including shape-based and ground-truth-based metrics, it is observed that the OPTICS algorithm fed by dual-stream input streams outperformed the rest of the investigated clustering algorithms and input sets. OPTICS algorithm groups load events based on their density in multi-dimensional space, using a dynamic radius. The OPTICS algorithm, as the best-performing transient state clustering algorithm for the low-frequency NILM purpose, is then tested with the downsampled input data in a wide frequency range, to observe the impact of the data-sampling frequency on the results, which simplifies the use of clustering algorithms in future studies.
引用
收藏
页数:9
相关论文
共 21 条
[1]  
Anderson K, 2012, POLICY REV, P35
[2]  
Anderson KD, 2012, IEEE IND ELEC, P3312, DOI 10.1109/IECON.2012.6389367
[3]   Evaluation of classification methodologies and Features selection from smart meter data [J].
Azaza, Maher ;
Wallin, Fredrik .
PROCEEDINGS OF THE 9TH INTERNATIONAL CONFERENCE ON APPLIED ENERGY, 2017, 142 :2250-2256
[4]  
Batra N, 2013, P 5 ACM WORKSH EMB S, P1
[5]  
Bernard T, 2018, INT CONF SMART GR C, P174, DOI 10.1109/ICSGCE.2018.8556735
[6]   A New Measurement Method for Power Signatures of Nonintrusive Demand Monitoring and Load Identification [J].
Chang, Hsueh-Hsien ;
Chen, Kun-Long ;
Tsai, Yuan-Pin ;
Lee, Wei-Jen .
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, 2012, 48 (02) :764-771
[7]   A Novel Approach Based on Finite-State Machines with Fuzzy Transitions for Nonintrusive Home Appliance Monitoring [J].
Ducange, Pietro ;
Marcelloni, Francesco ;
Antonelli, Michela .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2014, 10 (02) :1185-1197
[8]  
Gupta S, 2010, UBICOMP 2010: PROCEEDINGS OF THE 2010 ACM CONFERENCE ON UBIQUITOUS COMPUTING, P139
[9]   NONINTRUSIVE APPLIANCE LOAD MONITORING [J].
HART, GW .
PROCEEDINGS OF THE IEEE, 1992, 80 (12) :1870-1891
[10]   Applications of NILM Techniques to Energy Management and Assisted Living [J].
Hernandez, A. ;
Ruano, A. ;
Urena, J. ;
Ruano, M. G. ;
Garcia, J. J. .
IFAC PAPERSONLINE, 2019, 52 (11) :164-171