Postfault optimal islanding of smart grids using a reinforcement learning approach
Authors:
Open Access
- 15 March 2023
- journal article
- research article
- Published by Institution of Engineering and Technology (IET) in IET Generation, Transmission & Distribution
- Vol. 17 (11) , 2471-2482
Abstract
This paper presents a method for optimal reconfiguration of smart grids following the occurrence of short circuit faults. Due to restoration delays, the aim of the proposed approach is to save the maximum possible number of loads by forming stable islands and serving loads with Distributed Energy Resources (DERs). The islanding plan aims to prevent island instability and to help DERs continue supplying the maximum number of loads by the optimal network reconfiguration. Fault isolation is carried out by the protection system and the proposed procedure is commenced right after the fault isolation by controlling the condition of the network remote‐controlled switches. The proposed islanding plan is a novel method by this paper in the management of the postfault conditions of smart grids. Furthermore, a Q‐learning reinforcement approach is presented as the optimization tool because of its great capability and fast response for the determination of optimal reconfiguration. Numerous simulation tests for various fault locations on a 6‐bus and a 33‐bus test networks show the effectiveness of the proposed method in the improvement of postfault network reliability and sustainability.Keywords
This publication has 30 references indexed in Scilit:
- Autonomous microgrid operation subsequent to an anti-islanding schemeSustainable Cities and Society, 2018
- Non-communication protection method for meshed and radial distribution networks with synchronous-based DGInternational Journal of Electrical Power & Energy Systems, 2017
- Implementation of a new remote islanding detection method for wind–solar hybrid power plantsRenewable and Sustainable Energy Reviews, 2016
- A remote islanding detection and control strategy for photovoltaic-based distributed generation systemsEnergy Conversion and Management, 2015
- An Adaptive Overcurrent Protection Scheme for Distribution NetworksIEEE Transactions on Power Delivery, 2014
- A novel neural network and backtracking based protection coordination scheme for distribution system with distributed generationInternational Journal of Electrical Power & Energy Systems, 2012
- Simultaneous allocation of DGs and remote controllable switches in distribution networks considering multilevel load modelInternational Journal of Electrical Power & Energy Systems, 2011
- Reinforcement Learning approaches to Economic Dispatch problemInternational Journal of Electrical Power & Energy Systems, 2011
- A Reinforcement Learning algorithm to economic dispatch considering transmission losses2008
- Micro-Grid Autonomous Operation During and Subsequent to Islanding ProcessIEEE Transactions on Power Delivery, 2005