Least square estimator and IEC-60891 procedure for parameters estimation of single-diode model of photovoltaic generator at standard test conditions (STC)
Like some previous works, it is shown that ( I – V ) curve analysis can be applied to photovoltaic (PV) modules and to PV plants. Generally, manufacturers give data at standard test conditions (STC); therefore, for performance analysis and diagnosis of PV modules it is useful to extrapolate data and...
Gespeichert in:
Veröffentlicht in: | Electrical engineering 2021-04, Vol.103 (2), p.1253-1264 |
---|---|
Hauptverfasser: | , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Like some previous works, it is shown that (
I
–
V
) curve analysis can be applied to photovoltaic (PV) modules and to PV plants. Generally, manufacturers give data at standard test conditions (STC); therefore, for performance analysis and diagnosis of PV modules it is useful to extrapolate data and parameters at STC according to the International Electrotechnical Commission (IEC)-60891 procedures. Review on modeling PV generator is first established in this paper; next, the combination of the optimization method of least squares estimator algorithm and Newton–Raphson (NR) resolution for identifying the unknown parameters of single diode photovoltaic generators at STC is proposed; if parameters and curves are not obtained at STC, they are then extrapolated at STC according to IEC-60891 procedures. For polycrystalline silicon MSX60 PV generators, the dynamic variations of parameter values are carried out by graphs and the (
I
–
V
) curves are superposed, justifying the accuracy of the proposed method. The method is then extended to a 500 kW PV generator, the obtained parameters are then extrapolated to STC according to IEC-60891. The (
I
–
V
) and (
P
–
V
) curves obtained by extrapolated parameters are then compared with those measured. After up to 222 iterations, the parameters converge to constant values. Simulation shows maximum error value of 0.06 between experimental curve and estimated curves. |
---|---|
ISSN: | 0948-7921 1432-0487 |
DOI: | 10.1007/s00202-020-01131-2 |