An Adaptive Amplitude-Modulated Pseudo-Random Binary Sequence Excitation for Converter-Based Impedance Spectroscopy Characterization of Photovoltaic Modules

Abstract

Pseudo-random binary sequence (PRBS) has been implemented through DC-DC converters for internal impedance identification and shown to be faster and easier to implement than analogue multiple frequency signals; this is because a PRBS is a digitally generated binary signal, thus has faster transients, and requires a few samples for excitation signal reconstruction. PRBS however, is strictly designed for linear excitation, but because both converter and PV module being non-linear and operated in their pseudo-linear range for system identification, presents linearity issues in induced excitation systems. As the excitation frequency and module operating point vary, the induced excitation signal also varies, which may push the system into its non-linear range of operation – this can cause significant inaccuracies in the impedance measurements. This study thus, demonstrates the use of an adaptive, two-level optimized APRBS parametric excitation signal, designed by defining an input space, for the signal design points, characterized by the dynamic system from the converter's APRBS modulating duty ratio signal to the induced APRBS excitation signal. This will mitigate frequency and variable operating point induced system non-linearities, while producing fast and accurate online impedance estimation. The results show that the APRBS achieves good tolerance for the induced system non-linearities – this is demonstrated by a small impedance error between the designed system and a Frequency Response Analyzer (FRA) at low frequencies compared to PRBS excitation.