Resumen

ORDONEZ PALACIOS, LUIS EDUARDO; BUCHELI GUERRERO, VÍCTOR  y  CAICEDO BRAVO, EDUARDO. E-solar: a tool for solar resource assessment based on a Big Data architecture in a PySpark environment. Ing. Desarro. [online]. 2025, vol.43, n.1, pp.6-23.  Epub 03-Ene-2025. ISSN 0122-3461.  https://doi.org/10.14482/inde.43.01.456.089.

Over time, diverse researchers have created mathematical, statistical, and predictive models to evaluate solar resources. However, their implementation in technical tools restricts their usability for non-technical users. Additionally, data processing to estimate solar radiation often necessitates powerful hardware. This study introduces a Big Data based tool that employs flat files and satellite images to estimate solar radiation in Colombia. A model was developed using machine learning techniques and various programming languages. It operates within MapR, a distribution of the Hadoop ecosystem with an extensive array of Big Data capabilities and utilizes the PySpark API for parallel data processing within a computer cluster. The E-Solar tool, deployed on a web server, underwent assessment by professionals within the energy sector. Usability was analyzed, compliance with recent programming standards was confirmed, and profiles of interested users were identified. The solar radiation data generated by the tool are pivotal for solar projects. Furthermore, the tool lends support to researchers and organizations in decision-making for the implementation of photovoltaic systems, as it offers pertinent information regarding the behavior of solar resources in Colombia.

Palabras clave : Big Data; machine learning; MapR; PySpark; solar radiation.

        · resumen en Español     · texto en Español     · Español ( pdf )