NOVEL MPPT ALGORITHM BASED ON INDIVIDUAL PV PANEL MONITORING SYSTEM P. Guerriero (1) , S. Daliento (1) , M.Gargiulo (1) , V. d'Alessandro (1) , L. Farese (2) (1) University of Naples “Federico II” Dept. Electronic Engineering, Via Claudio 21, 80125 – Naples, Italy Phone: +39 081 7683126 Fax: +39 081 5934448 e-mail: pierluigi.guerriero@unina.it (2) I.S.E.T. Energia s.r.l., via Votta, Consorzio ARCHO, 81020, Valle di Maddaloni (CE), Italy Phone: +39 0823 23015 Fax: +39 0823 403752 e-mail: luca.farese@isetsrl.it ABSTRACT: The management of the power provided by a PV system can be effectively improved if local information about the operation of single panels are available. In this work a novel Maximum Power Point Tracking (MPPT) algorithm that exploits the knowledge of the actual operating condition of each solar panel of the PV array to track the maximum power point of the array independently of possible mismatching, e.g., caused by local shadowing, among solar panels, is presented. In order to collect information about the operating point of a solar panel connected in series with other panels, as part of a PV array system, a distributed wireless sensor network is proposed. A reduction of more than 50% in convergence time (when mismatching occurs) was observed by applying the new algorithm. Keywords: MPPT, PV Array, Shading, Mismatch. 1 INTRODUCTION The global efficiency of a grid-connected PV system is noticeable lower than the efficiency showed by individual cells or panels. This is largely due to mismatches among PV panels of the solar field often caused by temporary shadowing. The circuits devoted to the managing of the power delivered by the field (i.e. inverters implementing Maximum Power Point Tracking algorithms) only see, at their input terminals, the global behaviour of the field itself so that their performances are affected by the not-ideal shape of the IV curve they receive. In the recent literature attempts to overcome this problem have been presented; they are based on a distributed approaches where each PV panel is equipped with its own DC-DC circuit, performing a local maximum power tracking, so that each panel operates in its own [1]. This approach is, obviously, extremely onerous and risky because multiplies for the number of panels the probability of malfunction. In this work we propose an intermediate approach based on the idea that, if local information about the actual operation of each PV panel are available, MPPT algorithms can become highly effective because the best operating condition of the solar field can be computed off-line so that local maxima problems are completely avoided and convergence time is drastically reduced (ideally cancelled). To this end we developed a monitoring circuit that, applied to each panel, measures its IV performances during normal operation. Measurements are transmitted, via wireless, to a central unit interfaced with the inverter; information thus collected allow to evaluate the position of the maximum power point of the PV field and set the starting point of the MPPT algorithm directly at the maximum power condition. 2 THE REMOTE CIRCUIT The sense circuit, whose photo is shown in Fig.1, is composed of five different sections. The power supply section assures that the circuit is fully self-powered. It is based on super-C devices directly charged by the solar panel. The sense section provides direct measurements of the operating point of the PV panel. In particular, the open-circuit voltage Voc and the short-circuit current Isc are measured at arbitrary time intervals during the day. Each measurement takes less than 10 ms so that the string operation is absolutely unperturbed and power consumption is negligible (about 100 mW). The logic section manages timing and data flow. It is based on a 8-bit low power microcontroller that also allows storing and processing of measurements. The wireless communication section is based on the MiWi protocol (IEEE 802.15.4 compliant, 2.4 GHz frequency). The disconnecting section assures disconnection from the string during measurements [2-3]. Figure 1: Monitoring circuit PCB prototype. 3 THE ALGORITHM Classical MPPT algorithms are prone to local maxima convergence because they only work on integral information available at the field terminals; in particular they “look for” the maximum on the power vs voltage characteristic. Our approach exploits the knowledge of the operation status of each panel to directly set the algorithm on the 25th European Photovoltaic Solar Energy Conference and Exhibition / 5th World Conference on Photovoltaic Energy Conversion, 6-10 September 2010, Valencia, Spain 4451