Work

Multilayer feed-forward neural network-based mobile app for photovoltaic module parameter estimation

Year: 2026
Type: article
Abstract: Accurate parameter estimation of photovoltaic (PV) cells and modules is essential for the modeling, simulation, and optimization of PV systems. However, this task involves solving a complex nonlinear optimization problem, which often suffers from slow convergence and entrapment in local minima. The objective of this research is to contribute finding an efficient and software-independent approach to predict the parameters of the single diode model for polycrystalline and monocrystalline PV cells and modules from only the datasheet parameters at Standard Test Conditions (STC). To accomplish this task, an Artificial Neural Network (ANN) approach is used, which resulted in the ability to estimate instantly without the need of commercial software. The developed model is additionally implemented in an Android application, which offers a user friendly and accessible instrument for PV system designers. The model was further evaluated through comparisons of reconstructed I–V and P–V characteristics with available reference characteristics, Simulink library modules, manufacturer datasheet values, and recently published algorithms. The results demonstrate high reliability, with RMSE, MAE and U95 values of 1.3610e−04, 6.8160e−05 and 2.659820e−04 respectively. Furthermore, the model reliably reproduced P–V and I–V characteristics under varying environmental conditions, closely aligning with the datasheet specifications of the four modules.
Language: English
Field: Energy
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7233589814
FWCI:
Cites: 30
Cited by:
Related to: