Skip to main navigation Skip to search Skip to main content

Reinforcement learning for the unit commitment problem

Gal Dalal, Shie Mannor

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

Abstract

In this work we solve the day-ahead unit commitment (UC) problem, by formulating it as a Markov decision process (MDP) and finding a low-cost policy for generation scheduling. We present two reinforcement learning algorithms, and devise a third one. We compare our results to previous work that uses simulated annealing (SA), and show a 27% improvement in operation costs, with running time of 2.5 minutes (compared to 2.5 hours of existing state-of-the-art).

Original languageEnglish
Title of host publication2015 IEEE Eindhoven PowerTech, PowerTech 2015
ISBN (Electronic)9781479976935
DOIs
StatePublished - 31 Aug 2015
EventIEEE Eindhoven PowerTech, PowerTech 2015 - Eindhoven, Netherlands
Duration: 29 Jun 20152 Jul 2015

Publication series

Name2015 IEEE Eindhoven PowerTech, PowerTech 2015

Conference

ConferenceIEEE Eindhoven PowerTech, PowerTech 2015
Country/TerritoryNetherlands
CityEindhoven
Period29/06/152/07/15

Keywords

  • Learning (artificial intelligence)
  • Optimal scheduling
  • Optimization methods
  • Power generation dispatch

All Science Journal Classification (ASJC) codes

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering
  • Renewable Energy, Sustainability and the Environment

Fingerprint

Dive into the research topics of 'Reinforcement learning for the unit commitment problem'. Together they form a unique fingerprint.

Cite this