1551-3203 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TII.2015.2496202, IEEE Transactions on Industrial Informatics 1 Abstract-- Smart Grids (SG) are essential for efficient management and monitoring of electric power networks. One of the most important tasks in SG focuses on fault detection and automatic network reconfiguration. This process allows minimizing power losses and load balancing in distribution networks. In this paper an adaptation of the Open Shortest Path First (OSPF) routing protocol to accomplish the network reconfiguration task, is proposed. The algorithm is intended to run in secondary substation nodes over an agent-based distributed architecture. The proposed algorithm has been tested on IEEE 123 modified node test feeder and on an actual grid, deployed by an electrical distribution company. Moreover, a performance comparison with a typical centralized reconfiguration algorithm is carried out. Index Terms-- distributed approach, multiagent sytems, network reconfiguration, OSPF, smart grids. I. INTRODUCTION HE SG represent the future solution for electricity transmission and distribution networks [1]-[3] as they aim to achieve the complete monitoring of the network, allowing a better power balance between producers and consumers. SG systems must be able to react quickly and predictably, adapting to changing energy and supply demand, by controlling energy consumption and energy storage devices. The applications that manage the SG must rely on a secure communications network, highly scalable and always available. In addition, the communications infrastructure of the SG must support a large amount of real time data in order to react to changes of state, while providing Quality of Service (QoS). The distribution networks reconfiguration brings a chance to the system to re-routing the power after a fault occurrence, or to optimize some given criteria. The second case is especially interesting in grids with distributed generation resources, in which the reconfiguration can be implemented in order to optimize reliability, power loss or other electrical criteria. Copyright © 2009 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained from the IEEE by sending a request to pubs-permissions@ieee.org. Francisco J. Rodríguez, Susel Fernández, Inés. Sanz,, Miguel. Moranchel and Emilio J. Bueno are with the Group of Electronic Engineering Applied to Renewable Energy Systems in Department of Electronics, University of Alcalá, Spain (fjrs@depeca.uah.es; susel.fernandez@uah.es; ines.sanz@depeca.uah.es; miguel.moranchel@depeca.uah.es; emilio@depeca.uah.es). Typically, most of reconfiguration algorithms work in a centralized way and are executed in a SCADA or other primary substation software. However, since the number of secondary substation with monitoring equipment is growing, according with the SG deployment, it is feasible to use reconfiguration algorithms implemented and executed in a distributed manner. So, in this work a reconfiguration method is proposed, using a distributed approach, considering the reliability and power losses as optimization criteria. The proposal is to adapt a communication protocol (the OSPF routing protocol) for use in reconfiguring distribution networks. The algorithm have been implemented using an embedded multiagent system (MAS) deployed in secondary substations and the results are compared with those of Particle Swarm Optimization (PSO) algorithm which represents a centralized approach running in a primary substation. MAS are specially designed for networking, with a horizontal and collaborative decision-making level. The use of MAS applied to SG has been described in many publications as [4]-[11]. The rest of the paper is organized as follows. Section II contains an overview of reconfiguration algorithms. In section III the proposed multiagent architecture is presented. The reconfiguration algorithm proposed is explained in section IV. In section V the experimental setup and results can be found, and finally the conclusions and future work are drawn in section VI. II. OVERVIEW OF RECONFIGURATION ALGORITHMS Grid reconfiguration is ultimately a discrete non-linear optimization problem. As it can be considered like an optimization problem, there are some optimization algorithms that can be used to solve the problem: the Ant Colony Optimization (ACO), the PSO, the Evolutionary Algorithms, the Genetic Algorithms, the Fuzzy Control, etc. In the technical literature there are some examples of optimization algorithms applied to the grid reconfiguration problem. In [12] a hybrid evolutionary optimization algorithm based on combining ACO and Simulated Annealing is proposed for distribution feeder reconfiguration, considering distributed generators. In [13] ACO algorithm is used for grid reconfiguration according to the following criteria: reduction of power loss, balancing loads, service restoration after a fault and minimization of voltage deviation. Distributed Approach for SmartGrids Reconfiguration based on the OSPF routing protocol Francisco J. Rodriguez, Member IEEE, Susel Fernandez, Ines Sanz, Student Member IEEE , Miguel Moranchel, Student Member IEEE and Emilio J. Bueno, Member IEEE T