ORIGINAL CONTRIBUTION A Novel Control Strategy for Autonomous Operation of Isolated Microgrid with Prioritized Loads R. Hari Kumar 1 • S. Ushakumari 1 Received: 4 August 2015 / Accepted: 27 April 2018 Ó The Institution of Engineers (India) 2018 Abstract Maintenance of power balance between genera- tion and demand is one of the most critical requirements for the stable operation of a power system network. To miti- gate the power imbalance during the occurrence of any disturbance in the system, fast acting algorithms are inevitable. This paper proposes a novel algorithm for load shedding and network reconfiguration in an isolated microgrid with prioritized loads and multiple islands, which will help to quickly restore the system in the event of a fault. The performance of the proposed algorithm is enhanced using genetic algorithm and its effectiveness is illustrated with simulation results on modified Consortium for Electric Reliability Technology Solutions (CERTS) microgrid. Keywords Microgrid  Load shedding  Reconfiguration  Genetic Algorithm  Prioritized Loads Introduction Environmental concerns and increasing consumer demand for reliable high quality power stimulated the demand for restructuring the conventional power system network. Microgrid, which is a local grid with distributed generators, storage and loads is one of the proposed solutions to address these issues. Besides this, increased thrust for the development of microgrid in the past decade is mainly due to the rapid increase in the penetration of Distributed Generators (DGs), mostly renewable based, into the sys- tem. Microgrids may operate in grid connected mode or in isolated mode. In grid connected mode, the microgrid is connected to the main grid and it will act as a subsystem of the main grid. This subsystem will absorb power from the main grid when there is a shortage of power and will supply power to the main grid when there is excess power. Hence reliability is not a major issue in the grid connected mode. In isolated mode, the demand is to be met from the renewable energy sources and other DGs within the grid. Due to the intermittent nature of the renewable energy supply from various sources and the capacity limitation of the DGs, there are definite possibilities of power mismatch in this mode of operation. Also, when a fault occurs in the system, the deviation between generation and demand will be usually large. Hence, proper and efficient management of isolated microgrid, which includes generation rescheduling, load shedding and reconfiguration of the topology is absolutely necessary to mitigate the gap between demand and generation. This necessitates an intelligent and fast acting algorithm for load shedding and reconfiguration. Reconfiguration of conventional distribution system has attracted the attention of researchers since past few decades [1]. Soft computing techniques such as neural network [2, 3], Genetic Algorithm (GA) [4–6], simulated annealing [7, 8], Particle Swarm Optimization (PSO) [9, 10], evolu- tionary programming [11, 12] and ant colony optimization [13, 14] are extensively used in the reconfiguration of distribution system. The main objectives in these studies are to improve the voltage profile and to minimize the system losses. However, as the microgrid network exists in a small area, the loss minimization is not an objective for reconfiguration. Hence, the algorithms used for & R. Hari Kumar harikumar@cet.ac.in 1 Department of Electrical Engineering, College of Engineering Trivandrum, Thiruvananthapuram, Kerala, India 123 J. Inst. Eng. India Ser. B https://doi.org/10.1007/s40031-018-0335-7