2019 6th International Conference on Control, Decision and Information Technologies (CoDIT) | 2019
Innovative Single Sensor Neural Network PV MPPT
Abstract
This paper concerns the development of a single sensor neural network MPPT controller used to track the maximum power of PV generator system. The proposed MPPT controller has been developed in two steps: firstly, the proposed single sensor NN MPPT is trained in offline mode for testing different set of neural network parameters and architectures; secondly, the developed single sensor NN MPPT has been used in online mode to track the output power of the PV system composed of Solarex MSX-60W PV panel powering a resistive load via a DC-DC boost converter controlled using the proposed MPPT. The single sensor NN MPPT has been implemented and compared to the classical single sensor MPPT using Matlab/Simulink. Simulation results show that the proposed single sensor NN MPPT outperforms the classical single sensor MPPT in dynamic as well in static performances. In addition, the proposed single sensor NN MPPT controller can effectively track the maximum power using only one sensor compared to the classical MPPTs using two sensors reducing by the way the cost and the complexity of the PV MPPT controller.