Hybrid ANN-Linear Controller for Maximum PV Energy Harvesting in Grid-Tied Packed E-Cell Inverter

This paper proposes a novel auxiliary control loop based on irradiation and ambient temperature for online estimation of the inverter current; consequently, Maximum Power Point Tracking (MPPT) is achieved while Photo Voltaic (PV) panel is connected to the grid through the Nine-Level Packed E-Cell (PEC9) inverter. To this end, a hybrid control method including an Artificial Neural Network (ANN), PID and PI controllers is designed to achieve the Maximum Power (MP) of PV system connected to grid using Nine-Level PEC9 as a cost-effective and high efficient inverter topology. MPPT is performed by a boost converter controlled by Perturbation and Observation (P&O) algorithm. PEC9 is controlled by a Multilayer Perceptron (MLP) which increases the closed loop robustness performance against the parameters variations. The experimental and simulation studies have been implemented by dSpace-1104, MATLAB software and JKM265P-60 PV panel. The experimental results show that PEC9 inverter succeeds in injecting the MP of PV to the grid with least THD level even in shading and variable temperature conditions.