Abstract

The perturbation and observation (P&O) or hill-climbing maximum power point tracking (MPPT) algorithms are commonly used in PV systems due to their easy implementation. A P&O algorithm based on peak current control (PCC) and on the use of instantaneous sampled values to calculate the next perturbation can provide faster transient responses and small oscillations around the maximum power point (MPP) than when pulse width modulation (PWM) and averaged control values are used. However, the use of a fixed size perturbation (variation of the reference current for the PCC) results in a compromise solution between transient and steady-state responses. This thesis focuses on alternatives for implementing variable size perturbations in peak current controlled P&O MPPTs. First a Fuzzy logic based implementation is proposed and designed. Then hybrid region-based methods, where the MPPT algorithms operate differently depending on where in the PV panel Volt-current characteristics the system operates, are considered. The concept of non-switching zones is proposed as a means for moving the operating point of the PV system towards the vicinity of the MPP in the shortest possible interval. The potential performance of four different P&O algorithms is investigated by means of simulations. Experimental results are then used to verify how the computational burden of each algorithm and the processing speed of a common digital signal processor (DSP) affect the performance of each method in a practical prototype.