Design and Implementation of a Fuzzy Logic Based Maximum Power Point Tracker for a Photovoltaic System

Abstract

This thesis presents a method used to optimize the energy extraction in aphotovoltaic (PV) power system. The maximum power of a PV module changeswith temperature, solar radiation, and load. To increase efficiency, PV systems usea Maximum Power Point Tracker (MPPT) to continuously extract the highestpossible power and deliver it to the load. An MPPT consists of a dc-dc converterand a controller. The MPPT finds and maintains operation at the maximum powerpoint using a tracking algorithm. Many such algorithms have been proposed.However, the existing methods have drawbacks in terms of efficiency, accuracy,and flexibility. Due to the nonlinear behaviour of PV module current-voltagecharacteristics and the nonlinearity of converters due to switching, conventionalcontrollers are unable to give a good response in the presence of wide parametervariations and line transients.The objective of this research was to design and implement an MPPT thatuses a fuzzy logic control algorithm. Fuzzy logic, by dealing naturally withnonlinearities, offers a superior controller for this type of application. Thetechnique also benefits from the heuristic approach to the problem that overcomesthe complexity in modeling nonlinear systems. In order to achieve this goal, anMPPT model consisting of a PV module, a dc-dc converter, and a fuzzy logiccontroller was developed. Analysis of buck, boost, and buck-boost converter- xxiv -characteristics was carried out in order to identify the most suitable topology. Anintegrated model of the PV module and the identified converter was simulated andthe results used to derive the expert knowledge needed to formulate and tune thefuzzy logic controller. The controller was coded as a real-time control program andthe MPPT implemented using a dc-dc converter controlled by a microcomputer.The proposed method shows improved performance in terms of oscillationsabout the maximum power point, speed, and sensitivity to parameter variation. Theresults indicate that a significant amount of additional energy can be extracted froma photovoltaic module by using a fuzzy logic based maximum power point tracker.This results in improved efficiency for the operation of a photovoltaic power systemsince batteries can be sufficiently charged and used during periods of low solarradiation. The improved efficiency is expected to lead to significant cost savings inthe long run.