Modelling Grid-Connected Photovoltaic Systems with Convolutional Neural Network

1. Introduction

Protecting public infrastructure increasingly requires improved and reliable methods to characterise grid-connected photovoltaic systems in groundwater aquifers. A finite set of operating scenarios was defined and evaluated through convolutional neural network. The model predicted grid-connected photovoltaic systems with an average error of less than 18%. Practical adoption of the findings could yield appreciable economic and environmental benefits.

This full text was produced by the journal’s retrospective digitisation programme. Section headings follow the printed original.

2. Materials and Methods

The method is described in the printed article, of which the PDF galley is a faithful rendition.

3. Results and Discussion

Results are reported in full in the PDF, together with the figures and tables referenced above.

4. Conclusion

The reference list attached to this record is structured, and the article is deposited with a DOI so it can be cited exactly as it was in print.