Contents lists available at ScienceDirect Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes Distributed population dynamics for active and reactive power dispatch in islanded microgrids ☆ Nohora España a , John Barco-Jiménez b,c, ⁎ , Andrés Pantoja d , Nicanor Quijano e a Departamento de Ingeniería Electrónica, Corporación Universitaria Autónoma de Nariño, Pasto, Colombia b Programa de Posgrado en Ingeniería Eléctrica y Electrónica - PPIEE, Universidad del Valle, Cali, Colombia c Programa de Ingeniería Electrónica, Universidad CESMAG, Pasto, Colombia d Departamento de Electrónica, Universidad de Nariño, Pasto, Colombia e Departamento de Ingeniería Eléctrica y Electrónica, Universidad de los Andes, Bogotá, Colombia ARTICLE INFO Keywords: Economic dispatch Hierarchical control Reactive compensation Replicator dynamics ABSTRACT This work proposes a distributed strategy to solve a joint active and reactive power dispatch in isolated mi- crogrids. The information shared among neighboring local controllers allows the algorithm to achieve the op- timal dispatch in a secondary layer, which is the reference for primary controllers in each generator. The method based on population dynamics facilitates the implementation in systems with high penetration of distributed generation and a basic communication network. The hierarchical strategy is evaluated in a co-simulation platform to emulate real conditions in a study case of a microgrid with a management control scheme pro- grammed with the distributed optimization technique. Results show that the proposed technique provides an optimal dispatch in distinct scenarios with the expected reduction of losses, improvement in voltage profiles, and minimization of the generation costs. 1. Introduction The increasing participation of distributed generators (DGs) in the recent energy systems can solve some feeder-congestion and voltage- profile issues in the distribution networks. Microgrids, as small-scale distribution networks composed of DGs, storage devices, and inter- connected loads, are systems whose hierarchical control structure fa- cilitates the optimal coordination of the generation units [1]. In con- sequence, microgrids represent an efficient model to integrate renewable sources to traditional grids, reduce the greenhouse gas emissions, and provide energy to remote off-grid populations by means of island-operation solutions [2], which has led to a large deployement of microgrids in many countries [3]. However, this penetration requires coordination and control strategies such that the generation units may also take part in the frequency, voltage, and costs regulation of the overall system. Inappropriate active and reactive power flows in distribution sys- tems contribute to increase losses, voltage drops, and grid instability. These problems are more evident in isolated microgrids where the system’s references (e.g, voltage and frequency set points) are not provided by the main grid [4]. Nevertheless, if the island-operation mode is programmed or produced by unexpected failures, an internal coordination system must establish the control parameters. Then, the management strategies should set the DGs through their local power- electronics interfaces to dispatch active and reactive power to ensure stability and guarantee permissible levels of frequency and voltage. Besides, due to the increased number of different DGs in microgrids, distributed strategies solve more efficiently the resource allocation in comparison with the centralized controllers. Although the microgrid control tasks can be assigned to a central controller [4], the networked system constraints motivate novel stra- tegies based on distributed processes to fasten the control loops by means of local information shares. Moreover, distributed techniques reduce the complexity of the communication networks and provide flexibility to structural changes due to the capacity of optimization with information provided by only a subset of agents in the system. Authors in [5–8] survey several distributed techniques to optimize distinct variables in hierarchical layers of the power systems. This hierarchical https://doi.org/10.1016/j.ijepes.2020.106407 Received 9 March 2020; Accepted 27 July 2020 ☆ This work has been supported in part by the Colombia Científica Program within Ecosistema Científico, Contract No. FP44842-218-2018, Vicerrectoría de Investigaciones - Universidad CESMAG, Acta 3-P18-2020, and Fundación CEIBA, Contrato No. 87067512. ⁎ Corresponding author. E-mail addresses: nohora.espana@aunar.edu.co (N. España), john.barco@correounivalle.edu.co, jebarco@unicesmag.edu.co (J. Barco-Jiménez), ad_pantoja@udenar.edu.co (A. Pantoja), nquijano@uniandes.edu.co (N. Quijano). Electrical Power and Energy Systems 125 (2021) 106407 0142-0615/ © 2020 Elsevier Ltd. All rights reserved. T