Skip to main navigation Skip to search Skip to main content

Explainable Machine Learning Framework for Site Suitability Assessment of Multi-Generation Renewable Energy Systems in Developing Regions

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Increasing deployment of utility-scale renewable energy systems for multi-generation power plants necessitates systematic and objective approaches for identifying appropriate geographic sites. This paper proposes an integrated framework of Machine Learning and Geographic Information System (MLGIS) techniques to assess the suitability of utility-scale solar and wind farms in developing regions. The method combines geospatial data preprocessing, supervised machine learning classification, explainable AI methods and a weighted overlay analysis to address the limitations related to multi-criteria decisionmaking methods. Datasets of presence-only data were created by combining coordinates of existing renewable energy installations with randomly generated pseudo-absence coordinates. Various machine learning algorithms, such as Logistic Regression, Random Forest, Gradient Boosting, and Multilayer Perceptron, were tested. Of these, Random Forest performed best in predicting suitable locations for solar and wind energy, achieving an area under the curve of about 0.95 and a classification accuracy of 91.5%. The technique of SHapley Additive exPlanations (SHAP) was employed to calculate criterion weights, which represented data-informed values incorporated in geographic information systems suitable location mapping. The results suggest that a major criterion determining suitable locations is proximity to existing electric power infrastructure and transportation networks, while solar intensity and wind speed function as secondary determinants. The suitability maps are compatible with existing renewable energy installations, indicating the applicability of the proposed framework for renewable energy planning and site selection.

Original languageEnglish
Title of host publicationSysCon 2026 - The 20th Annual IEEE International Systems Conference, Conference Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331591519
DOIs
StatePublished - 2026
Event20th Annual IEEE International Systems Conference, SysCon 2026 - Halifax, Canada
Duration: 6 Apr 2026 → 9 Apr 2026

Publication series

NameSysCon 2026 - The 20th Annual IEEE International Systems Conference, Conference Proceedings

Conference

Conference20th Annual IEEE International Systems Conference, SysCon 2026
Country/TerritoryCanada
CityHalifax
Period6/04/26 → 9/04/26

Bibliographical note

Publisher Copyright:
© 2026 IEEE.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

  1. SDG 7 - Affordable and Clean Energy
    SDG 7 Affordable and Clean Energy

Keywords

  • Explainable artificial intelligence (XAI)
  • GIS
  • Machine learning
  • Multigeneration energy systems
  • Renewable energy
  • Site suitability assessment

ASJC Scopus subject areas

  • Artificial Intelligence
  • Hardware and Architecture
  • Information Systems
  • Decision Sciences (miscellaneous)
  • Information Systems and Management
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Explainable Machine Learning Framework for Site Suitability Assessment of Multi-Generation Renewable Energy Systems in Developing Regions'. Together they form a unique fingerprint.

Cite this