AI Summary & Key Takeaways
AI InsightThese findings highlight that model selection should be guided by analytical objectives, with Mamdani preferred for qualitative and interpretive assessments, and Sugeno for precise numerical and computationally efficient analyses.
📌 3 Key Takeaways
Abstract
Risk assessment of gas transmission pipelines, particularly in mountainous regions with complex geomorphology, is essential for ensuring the safety and reliability of critical energy infrastructure. This study compares the Mamdani and Sugeno fuzzy inference systems for evaluating the risk of gas pipelines in Ilam city, Iran, by integrating fuzzy logic with Geographic Information Systems (GIS). Key risk factors—including distance from waterways, proximity to landslide-prone areas, pipeline leakage (dependent on material and diameter), service connection density, longitudinal expansion (dependent on material and length), and distance from gas valves—were identified and incorporated into the models. The final risk maps indicated that areas adjacent to watercourses, landslide-prone zones, regions with higher service connection density, and sections farther from control valves are at greater risk. Quantitative validation showed strong spatial agreement between the models, with IoU = 0.81 and Dice = 0.89, and high-risk areas covering 22.3% in the Mamdani model and 19.6% in the Sugeno model. The Sugeno model demonstrated 24% faster runtime and 18% lower memory usage, providing smoother numerical outputs suitable for quantitative analysis, whereas the Mamdani model offered clear, interpretable results advantageous for managerial decision-making. These findings highlight that model selection should be guided by analytical objectives, with Mamdani preferred for qualitative and interpretive assessments, and Sugeno for precise numerical and computationally efficient analyses.