A variable step-size MPPT for sensorless current model predictive control for photovoltaic systems

Variability of the solar energy resources requires highly effective maximum power point tracking (MPPT) to ensure maximum energy harvesting from the photovoltaic (PV) modules. To accomplish this, a MPPT controller typically requires accurate knowledge of the voltage and current from the PV module, and must converge quickly with minimal hunting around the maximum power point (MPP). Conventional MPPT techniques use fixed step-size perturbation which need to be optimized for one of two objectives: reducing the convergence settling time, or reducing the steady state ripple. Also, the required sensors increase system cost and can cause reliability issues, particularly for the current sensors which can exhibit thermal drift and degrade over time. This paper presents a highly efficient, variable-step sensorless current MPPT controller using an observer-based model derived from the principles of model predictive control (MPC) to adaptively determine the perturbation step-size. The proposed variable step, sensorless current, model predictive control maximum power point tracking (VS-SC-MPC-MPPT) continuously adjusts the perturbation step size using the predicted dynamic model to enable fast convergence and small limit cycle, without the need of expensive measuring devices. The performance of the VS-SC-MPC-MPPT in this paper is compared to previously developed MPC-MPPT methods. The provided investigation aims to demonstrate higher system efficacy with lower cost. The feasibility of the proposed controller is verified though computer simulation and real time simulation using dSPACE DS1007.