Evolutionary Optimization of Multi-Agent Control Strategies for Electric Vehicle Charging

Research output: Chapter in Book/Report/Conference proceedingsConference contributionpeer-review

11 Citations (Scopus)

Abstract

While an increasing share of intermittent and non-dispatchable renewable energy plants cause probabilistic behavior at the power grids' supply side, the expected penetration of electric mobility at the demand side offers the opportunity of controllable load. Their optimal coordination is one major concern for future smart grids. Therefore, a multi-agent system will be proposed where each electric vehicle (agent) acts in response to dynamic conditions in its environment according to a given strategy. Optimizing these strategies will be the core of this paper, while evolutionary computation will be used for optimization. Here, simulation models will be applied for problem representation and solution evaluation. Thus, simulation allows modeling of complex as well as probabilistic systems, necessary for the herein tackled probem. In the end, the optimized strategies determine electric vehicles' charging behavior such that end-users' energy demand is satisfied and secure power grid operation is guaranteed throughout the considered grid using power from renewable plants. For solution representation, two different approaches will be compared concerning reachable solution quality as well as problem-specific metrics.

Original languageEnglish
Title of host publicationGECCO'12 - Proceedings of the 14th International Conference on Genetic and Evolutionary Computation Companion
PublisherACM Sigevo
Pages3-10
Number of pages8
ISBN (Print)978-1-4503-1178-6
DOIs
Publication statusPublished - 2012
EventGenetic and Evolutionary Computation Conference GECCO 2012 - Philadelphia, United States
Duration: 7 Jul 201211 Jul 2012
http://www.sigevo.org/gecco-2012/

Publication series

NameGECCO'12 - Proceedings of the 14th International Conference on Genetic and Evolutionary Computation Companion

Conference

ConferenceGenetic and Evolutionary Computation Conference GECCO 2012
CountryUnited States
CityPhiladelphia
Period07.07.201211.07.2012
Internet address

Keywords

  • Electric vehicle charging control
  • Evolutionary strategies
  • Policy optimization
  • Simulation-based optimization

Fingerprint Dive into the research topics of 'Evolutionary Optimization of Multi-Agent Control Strategies for Electric Vehicle Charging'. Together they form a unique fingerprint.

Cite this