Nonlinear control design and stability analysis of hybrid grid-connected photovoltaic-Battery energy storage system with ANN-MPPT method

Abstract

The problem of controlling a grid-connected solar energy conversion system with battery energy storage is addressed in this work. The study’s target consists of a series and parallel combination of solar panel, DC/DC converter boost, DC/AC inverter, DC/DC converter buck-boost, Li-ion battery, and DC load. The main objectives of this work are: (i) PV voltage regulation: the PV panel voltage must track a given reference voltage corresponding to the maximum power point. The optimal reference voltage is generated by an optimizer based on the artificial neural network (ANN); (ii) DC link voltage regulation: The DC link voltage must track a reference voltage as closely as possible. (iii) PFC requirement: The grid currents must be sinusoidal with the same frequency and in phase with the voltage grid. The periodic nature of solar energy and the frequent fluctuations in load demand reduce battery life and charging performance. To deal with these limitations and ensure the battery’s safety, a battery energy management algorithm is developed with the following objectives: (iv) CC mode: When the battery’s state of charge is less than 100% (SOC<100%), it indicates that the battery is not fully charged, this aim has been adopted. The battery current should follow a constant reference as closely as possible to the permissible limits battery current; (v) CV mode: When the battery is completely charged, this mode is switched on. The voltage at the battery terminals then achieves its reference signal, which corresponds to the charge state’s of the battery SOC=100%. In addition, the energy management system which generates several energy flow scenarios, in this work, the focus is to balance the energy flows between the load and the different energy sources to minimize the system costs, to ensure the stability of the grid and to improve the power quality. To do this, a mathematical modeling of the overall system was performed. Subsequently, backstepping controllers are then synthesized in order to ensure the control objectives. The closed loop control convergence is formally analyzed using Lyapunov’s stability theory and its performances are illustrated by simulation. As a result, the simulation results indicate that the proposed controllers perform admirably in achieving their objectives.