Maximum power point traking controller for PV systems using neural networks

This paper presents a development and implementation of a PC-based maximum power point tracker (MPPT) for PV system using neural networks (NN). The system consists of a PV module via a MPPT supplying a dc motor that drives an air fan. The control algorithm is developed to use the artificial NN for detecting,the optimal operating point under different operating conditions, then the control action gives the driving signals to the MPPT. A PC is used for data acquisition, running the control algorithm, data storage, as well as data display and analysis. The system has been implemented and tested under various operating conditions. The experimental results showed that the PV system with MPPT always tracks the peak power point of the PV module under various operating conditions. The MPPT transmits about 97% of the actual maximum power generated by the PV module. The MPPT not only increases the power from the PV module to the load; but also maintains longer operating periods for the PV system. The air velocity and the air mass flow rate of the mechanical load are increased considerably, due to the increase of the PV system power. It is also found that, the increase in the output energy due to using the MPPT is about 45.2% for a clear sunny day. (c) 2004 Elsevier Ltd. All rights reserved.