Implementation of fuzzy-sliding mode based control of a grid connected photovoltaic system Abdelkrim Menadi a,n , Sabrina Abdeddaim b , Ahmed Ghamri b , Achour Betka b a Laboratoire LMSE, Electrical Engineering Department, University of Biskra, Algeria b Laboratoire LGEB, Electrical Engineering Department, University of Biskra, Algeria article info Article history: Received 7 January 2015 Received in revised form 26 May 2015 Accepted 11 June 2015 Available online 1 August 2015 This paper was recommended for publica- tion by Dr. Jeff Pieper Keywords: Photovoltaic Grid connected MPPT Sliding mode control Fuzzy logic Unity power factor abstract The present work describes an optimal operation of a small scale photovoltaic system connected to a micro-grid, based on both sliding mode and fuzzy logic control. Real time implementation is done through a dSPACE 1104 single board, controlling a boost chopper on the PV array side and a voltage source inverter (VSI) on the grid side. The sliding mode controller tracks permanently the maximum power of the PV array regardless of atmospheric condition variations, while The fuzzy logic controller (FLC) regulates the DC-link voltage, and ensures via current control of the VSI a quasi-total transit of the extracted PV power to the grid under a unity power factor operation. Simulation results, carried out via Matlab–Simulink package were approved through experiment, showing the effectiveness of the proposed control techniques. & 2015 Published by Elsevier Ltd. on behalf of ISA. 1. Introduction As conventional energy sources are vanishing fast with a conse- quent rise in cost, considerable attention is being paid to other alte- rnative sources. Nowadays, solar energy, which is free and abundant in most parts of the world, has proven to be an economical source of energy in many applications. Nowadays, grid connected PV systems have acquired a mature technology, and is receiving more attention in recent years as decentralized sources, to share the power demand in case of grid disturbances, improving therefore the stability of the interconnected networks. In this topic, various studies have been carried out on sizing [1,2], matching [3], and optimizing [4]. Different optimization strategies were proposed to improve the overall system efficiency, by extracting and then injecting the maximum PV power into the grid. The first task deals with tracking the maximum power point (MPP) of the PV array, where many algorithms are used; whereas, the second task consists in injecting this power with minimum losses. As exposed in [5,6], the conventional Perturb & Observe (PO) method does not provide a good accuracy and response time, since oscillation occurs around the optimal point in steady state [7]. To overcome this drawback, several intelligent and complex control methods, such as fuzzy logic [8,9] genetic algorithms [10,11], neural network [12], neuro-fuzzy MPPT strategies [13], were devel- oped in recent years to improve accuracy and response time. Compared to the classical algorithms, artificial techniques proved a notable superiority, since the maximum power point is always tracked very fast regardless any sudden changes of solar insolation, and without oscillation in steady state [14]. As an interface with the AC side, two or multi-level VSI with various pulse width modulation techniques, are currently in use. They are controlled either in current or voltage mode to allow the PV power flow of the PV power to the utility under a controlled power factor operation. In this scope, the present paper describes how an operation of a small scale PV system connected to a microgrid can be achieved. The main tasks assigned to the proposed control strategies are 1. A permanent tracking of the maximum power point of the PV array, by a proper tuning of the boost chopper duty cycle, using sliding mode based MPPT control. 2. A total flow of the extracted PV power to the utility, via current control of the VSI, under a unity power factor operation. In order to investigate the system performances, and prior to numerical simulation, each part of the system is modeled, taking into account the following assumptions: the synchronization of the PV system with the grid is not considered in this study, power converters Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/isatrans ISA Transactions http://dx.doi.org/10.1016/j.isatra.2015.06.009 0019-0578/& 2015 Published by Elsevier Ltd. on behalf of ISA. n Correspondence to: Laboratoire LGEB, Electrical Engineering Department, Uni- versity of Biskra, Algeria. Tel.: þ213 665150341. E-mail addresses: a.menadi@univ-biskra.dz (A. Menadi), s_abdeddaim@yahoo.fr (S. Abdeddaim), ghamri65@gmail.com (A. Ghamri), betkaachour@gmail.com (A. Betka). ISA Transactions 58 (2015) 586–594