Virtual Metering: An Efficient Water Disaggregation Algorithm via Nonintrusive Load Monitoring

被引:3
作者
Wang, Bingsheng [1 ]
Chen, Zhiqian [2 ]
Boedihardjo, Arnold P. [3 ]
Lu, Chang-Tien [4 ]
机构
[1] Google Inc, 1600 Amphitheatre Pkwy, Mountain View, CA 94403 USA
[2] Virginia Tech, Northern Virginia Ctr, Room 317,Haycock Rd, Falls Church, VA 22043 USA
[3] US Army, Corps Engineers, 4552 Fair Valley Dr, Fairfax, VA 22033 USA
[4] Virginia Tech, Northern Virginia Ctr, Room 310,Haycock Rd, Falls Church, VA 22043 USA
关键词
Computational sustainability; Bayesian discriminative learning; sparse coding; Mixture-of-Gammas; low-sampling-rate disaggregation; non-intrusive load monitoring; SPARSE; RECOGNITION; MODEL;
D O I
10.1145/3141770
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The scarcity of potable water is a critical challenge in many regions around the world. Previous studies have shown that knowledge of device-level water usage can lead to significant conservation. Although there is considerable interest in determining discriminative features via sparse coding for water disaggregation to separate whole-house consumption into its component appliances, existing methods lack a mechanism for fitting coefficient distributions and are thus unable to accurately discriminate parallel devices' consumption. This article proposes a Bayesian discriminative sparse coding model, referred to as Virtual Metering (VM), for this disaggregation task. Mixture-of-Gammas is employed for the prior distribution of coefficients, contributing two benefits: (i) guaranteeing the coefficients' sparseness and non-negativity, and (ii) capturing the distribution of active coefficients. The resulting method effectively adapts the bases to aggregated consumption to facilitate discriminative learning in the proposed model, and devices' shape features are formalized and incorporated into Bayesian sparse coding to direct the learning of basis functions. Compact Gibbs Sampling (CGS) is developed to accelerate the inference process by utilizing the sparse structure of coefficients. The empirical results obtained from applying the new model to large-scale real and synthetic datasets revealed that VM significantly outperformed the benchmark methods.
引用
收藏
页数:30
相关论文
共 55 条
  • [1] [Anonymous], 2011, PROC SIAM INT C DATA, DOI [10.1137/1, 10.1137/1.9781611972818.64, DOI 10.1137/1.9781611972818.64]
  • [2] [Anonymous], ARXIV161209106
  • [3] [Anonymous], 2010, P ADV NEUR INF PROC
  • [4] [Anonymous], INT C SPOK LANG PROC
  • [5] [Anonymous], 2011, P 17 ACM SIGKDD INT
  • [6] [Anonymous], 2009, Advances in Neural Information Processing Systems
  • [7] [Anonymous], 1993, SIGN SYST COMP 1993
  • [8] Independent factor analysis
    Attias, H
    [J]. NEURAL COMPUTATION, 1999, 11 (04) : 803 - 851
  • [9] Atomic decomposition by basis pursuit
    Chen, SSB
    Donoho, DL
    Saunders, MA
    [J]. SIAM JOURNAL ON SCIENTIFIC COMPUTING, 1998, 20 (01) : 33 - 61
  • [10] MAXIMUM LIKELIHOOD FROM INCOMPLETE DATA VIA EM ALGORITHM
    DEMPSTER, AP
    LAIRD, NM
    RUBIN, DB
    [J]. JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-METHODOLOGICAL, 1977, 39 (01): : 1 - 38