Dal 20 al 22 novembre 2024 si è tenuta in Sicilia l’ottava edizione dell’International Conference on System Reliability and Safety (ICSRS 2024) www.icsrs.org , un evento di riferimento per la comunità scientifica internazionale nel settore dell’affidabilità e della sicurezza dei sistemi. La conferenza è stata co-sponsorizzata dalla IEEE Reliability Society (Italy Chapter) e ha visto la partecipazione di esperti da tutto il mondo. Tra i temi affrontati: intelligenza artificiale e simulazione avanzata per all’affidabilità, sicurezza, resilienza dei sistemi complessi.

Ruoli chiave nella conferenza

Diversi docenti del del Laboratorio di Analisi di Sistemi per l’Analisi di Affidabilità, Rischio e Resilienza (LASAR³) del Dipartimento di Energia hanno avuto incarichi di rilievo nell’organizzazione dell’evento:

• Prof. Enrico Zio – Conference General Chair
• Prof. Piero Baraldi – Technical Program Chair
• Prof. Francesco Di Maio – Local Organizing Chair

Session Chair e Keynote Lecture

I docenti e ricercatori del Dipartimento hanno anche coordinato e presieduto diverse sessioni tematiche:

• Prof. Francesco Di Maio – Session Chair Risk Assessment 1
• Prof. Masoud Naseri – Session Chair Risk Assessment 2
• Dott. Ali Hosseini – Session Chair Safety Analysis in the Energy Sector
• Dott. Luca Pinciroli – Session Chair Methods for Anomaly Detection
• Dott. Dario Valcamonico – Session Chair Safety Analysis

Presentazioni orali e contributi scientifici

Tra gli interventi più rilevanti presentati dai membri del Dipartimento:

• Prof. Francesco Di Maio
  • “ARtificial Intelligence and STOchasTic Simulation for the rEsiLience of Critical InfrastructurES (ARISTOTELES)”
  • “Simulation-based Probabilistic Risk Assessment (SIMPRA) of Integrated Road-Power Infrastructures Travelled by EVs and ICVs”
• Prof. Ibrahim Ahmed – Oral Presentation:
  • “Critical Heat Flux Prediction by Physics-informed Neural Networks”
• Dott. Ali Hosseini – Oral Presentation:
  • “A Dynamic Bayesian Network for the Performance Assessment of Nuclear Waste Repositories Undergoing Chemical Degradation due to Climate Change”

• Dott. Luca Pinciroli – Oral Presentation:
  • “Anomaly Detection of Renewable Energy Systems by Unsupervised Graph Neural Networks”

Contributi dei dottorandi STEN

Oltre ai docenti e ricercatori del gruppo di ricerca LASAR³, gli studenti del corso di dottorato Energy and Nuclear Science and Technology – STEN del Dipartimento di Energia che hanno presentato i loro lavori scientifici sono:

• Mudi Jiang – Bayesian Deep Learning Framework with Variational Inference for Uncertainty Quantification in RUL Prediction

• Haoran Liu – Analysis of Three-dimensional (3-D) Warranty Data Considering Fluctuated Sales and Heterogeneous Failures

• Stefano Marchetti – A Deep Reinforcement Learning method for finding the risk-based optimal prescriptive maintenance policy of degrading safety barriers

• Giovanni Roma – Severe Accident Management of Nuclear Power Plants by A Condition-Informed Dynamic Bayesian Network

• Santiago Taguado Menza – Artificial Intelligence (AI) – Accelerated Stochastic Simulation for the Resilience Assessment of Critical Energy Infrastructures

• Zongyao Wang – A Sample Synthesis Method for Railway Track Circuit Fault Diagnosis Based on Variational Autoencoder and Improved SMOTEENN

• Weijun Xu – Uncertainty-Aware Prediction of Remaining Useful Life

• Yike Zhao – Deep Neural Network based on Domain Adaptation for the Prediction of Critical Parameters in Hazardous Plants

• Shiyu Chen – Production Planning Optimization of Energy Supply Chain based on the Agent-based Cooperative Framework

• Maria Valentina Clavijo Mesa – Inoperability Assessment of Interdependent Critical Infrastructures by Minimal Inoperability Sets Analysis

• Thomas Coscia – Probabilistic Safety Assessment of Nuclear Power Plants under Climate Change

• Mijie Du – Power Network Vulnerability Assessment Considering Spatial Heterogeneity Demand

• Giovanni Floreale – Improving the Performance of PHM Models by Post-hoc Explainable Artificial Intelligence (XAI) with Automatic Processing of Explanations

• Juan Pablo Futalef – A Classification Problem Formulation for the Reliability Assessment of End-To-End Communication among Cyber Devices of Cyber-Physical Systems