Cascade Neural Network Based MPPT Controller for Thermoelectric Generator

The thermoelectric generator (TEG) converts heat energy into electrical energy. The electrical power generated from TEG largely depends on the difference in the junction temperature of the TEG. The maximum power generated from the TEG varies when temperature difference varies. The efficient controller has to be designed to track the maximum power under varying temperature conditions so that the TEG always operate at the maximum power point for better efficiency. This paper proposes a novel neural network approach using cascade architecture for maximum power point tracking (MPPT) in TEG. The proposed method is compared with conventional feed-forward neural network based MPPT controller in terms of accuracy and complexity. The interesting and promising results obtained are presented.