Quantifying Flexibility of Residential Electric Vehicle Charging Loads using Non-Intrusive Load Extracting Algorithm in Demand Response

The increasing penetration of electric vehicles (EVs) has increased the operational burden on power grids during peak load periods. Identifying the charging patterns of residential EV clusters and quantifying their flexibility in demand response (DR) can help grid operators make effective regulatory decisions for balancing grid supply and load. In this study, statistical models are established for the charging patterns (charging start time, charging end time, and charge duration) of residential EVs. A training-free non-intrusive load extracting (NILE) algorithm is proposed. The algorithm is based on different load signatures and power block extremums, and can extract the EV load charged by different charging power. A price-based DR model is established to optimise the charging strategy of EVs, and the flexibility of EVs in DR is quantified according to the change of charging behaviour before and after optimisation. In addition, the validity of the NILE algorithm is verified with a real data set, and the flexibility of the EV cluster in different situations (seasons, weekdays, and weekends) is quantified based on the separated charging data.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01714176
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jun 24 2019 3:06PM