An efficient, fast, and robust algorithm for single diode model parameters estimation of photovoltaic solar cells
Husain A. Ismail,Ahmed A. Zaki Diab
DOI: https://doi.org/10.1049/rpg2.12958
IF: 3.034
2024-02-11
IET Renewable Power Generation
Abstract:The discrete search methodology can be highlighted through the following approaches: (a) Optimum discrete search: This approach focuses on identifying the optimal solution within a discrete set of values. It involves searching through a predefined set of discrete values to find the most suitable parameter estimation. (b) Proposed alternate optimization (AO) algorithm: The AO algorithm proposed in this paper is another aspect of the discrete search methodology. It leverages the discrete set of values to efficiently estimate the parameters in the single diode model circuit. The AO algorithm offers robust performance and faster convergence, further enhancing the accuracy of the parameter estimation process. Parameter estimation of photovoltaic (PV) solar cells and module models pays attention to researchers owing to their importance in practical considerations. The single diode model (SDM) circuit with five unknown parameters is widely used to model PV solar cells and modules. In this paper, a novel approach called alternate optimization (AO) algorithm based on a discrete search is proposed to estimate the SDM parameters. The proposed algorithm provides efficient and robust performance, considering a limited set of discrete values and increasing the convergence speed. Two practical case studies with actual measurements are considered to assess the proposed AO algorithm: the RTC France solar cell and monocrystalline PV modules with different irradiations and temperatures. The numerical findings underscore the superior performance of the proposed AO algorithm across various metrics. Notably, it achieves an exceptional Root Mean Square Error (RMSE) of 7.7426 × 10−04 for the RTC France PV cell and approximately 1 × 10−03 RMSE for monocrystalline PV modules. Additionally, the algorithm exhibits unparalleled speed, showcasing the fastest convergence with an elapsed time of 1.66 × 10−05—markedly 4.45 times quicker than the fastest method documented in the literature for SDM parameter estimation. Furthermore, the proposed AO algorithm stands out for its efficiency, requiring a maximum of five iterations for parameter estimation, a substantial improvement compared to the more than 10 iterations typically needed by algorithms in the existing literature. Its robustness is also commendable, as evidenced by the stability of final RMSE values across a variety of experiments, distinguishing it from less robust algorithms found in the literature.
energy & fuels,engineering, electrical & electronic,green & sustainable science & technology