Clustering-Based Residential Baseline Estimation: A Probabilistic Perspective

IEEE Transactions on Smart Grid - Tập 10 Số 6 - Trang 6014-6028 - 2019
Mingyang Sun1, Yi Wang2, Fei Teng1, Yujian Ye1, Goran Strbac1, Chongqing Kang2
1Department of Electrical and Electronic Engineering, Imperial College London, London, UK
2Department of Electrical Engineering, Tsinghua University, Beijing, China

Tóm tắt

Demand response (DR) is one of the most cost-effective solutions for providing flexibility to power systems. The extensive deployment of DR trials and the roll-out of smart meters enable the quantification of consumer responsiveness to price signals via baseline estimation. The traditional deterministic baseline estimation approach can provide only a single value without consideration of uncertainty. This paper proposes a novel probabilistic baseline estimation framework that consists of a daily load profile pool construction stage, a deep learning-based clustering stage, an optimal cluster selection stage, and a quantile regression forests model construction stage. In particular, the concept of a daily load profile pool is introduced, and a deep-learning-based clustering approach is employed to handle a large number of daily patterns to further improve the baseline estimation performance. Case studies have been conducted on fine-grained smart meter data collected from a real dynamic time-of-use tariffs trial of the low carbon London project. The superior performance of the proposed method is demonstrated based on a series of evaluation metrics regarding both deterministic and probabilistic estimation results.

Từ khóa

#Deep learning #demand response #probabilistic baseline estimation #clustering #dynamic time-of-use tariff

Tài liệu tham khảo

10.1126/science.1127647 xie, 2016, Unsupervised deep embedding for clustering analysis, Proc Int Conf Mach Learn, 478 tian, 2014, Learning deep representations for graph clustering, Proc AAAI, 1293 10.24963/ijcai.2017/243 tindemans, 2016, Low carbon London project: Data from the dynamic time-of-use electricity pricing trial, 2013 10.1109/TPWRS.2016.2614366 srivastava, 2014, Dropout: A simple way to prevent neural networks from overfitting, J Mach Learn Res, 15, 1929 10.1561/2200000006 10.2172/928452 10.1109/TPWRS.2015.2453889 10.3390/en80910239 10.1016/j.ijepes.2018.02.049 10.1109/PESGM.2015.7285756 10.1016/j.apenergy.2018.02.058 schofield, 2014, Residential consumer responsiveness to time-varying pricing 10.1016/j.ijforecast.2015.11.011 10.1109/TPWRS.2017.2688178 meinshausen, 2006, Quantile regression forests, J Mach Learn Res, 7, 983 wang, 0, Deep learning-based socio-demographic information identification from smart meter data, IEEE Trans Smart Grid 10.1038/s41560-018-0105-4 sun, 0, A deep learning-based feature extraction framework for system security assessment, IEEE Trans Smart Grid 10.1109/TSG.2015.2437451 10.1109/TSG.2018.2824842 10.1109/TSG.2017.2703842 wang, 0, Review of smart meter data analytics: Applications, methodologies, and challenges, IEEE Trans Smart Grid 10.1016/j.enbuild.2008.11.002 10.1109/TSG.2015.2463755 10.1016/j.epsr.2016.12.012 10.1109/TD-ASIA.2009.5356984 10.1016/j.apenergy.2017.06.066 10.1109/TIE.2018.2803732 10.1016/j.energy.2017.12.154 10.1109/TSG.2018.2807985 10.1186/s13673-018-0148-3 10.1016/j.energy.2011.12.031 10.1109/TSG.2017.2686012 10.1038/nature14539