• Blended Learning & Flipped Classroom

The course is offered in a 10-CFU version (ARTIFICIAL INTELLIGENCE AND ADVANCED SIMULATION FOR THE SAFETY, RELIABILITY AND MAINTENANCE OF ENERGY SYSTEMS) and in a two 5-CFU courses version (INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS and COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE). The following detailed program defines goals, learning outcomes and topics for each 5-CFU course.

INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS (5 ECTS)

The objective of the course is to provide students with an understanding of the Deterministic (D) and Probabilistic (P) Safety Analysis (SA) methods, and their integration (IDPSA), which requires Artificial Intelligence (AI) and advanced simulation methods. These will be presented with respect to practical cases of nuclear systems.

COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE (5 ECTS)

The goal of this course is to provide students with the knowledge on Artificial Intelligence (AI) and advanced simulation methods used for the reliability and availability analysis, and maintenance engineering of industrial equipment and energy systems.

INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS

The student knows:

• ELO IDPSA1 - the potential scenarios of nuclear accidents and the methods for their deterministic (DSA) and probabilistic analysis (PSA)
• ELO IDPSA2 - the methods for uncertainty and sensitivity analysis of nuclear accidents models and codes and the calculation of the safety margins

The student understands:

• ELO IDPSA3 – strengths and weaknesses of DSA and PSA, and the potential and challenges of IDPSA

The student is able to:

• ELO IDPSA4 - implement AI and simulation methods for DSA, PSA and IDPSA
• ELO IDPSA5 - apply AI and simulation methods for dynamic accidental scenarios analysis

The student can communicate:

• ELO IDPSA6 - the results of the application of AI and simulation methods to IDPSA

COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE (5 ECTS)

The student knows:

• ELO CMRAM1 - the basics of reliability and availability analysis, and maintenance engineering
• ELO CMRAM2 - AI and advanced simulation methods for fault detection, diagnostics and prognostics, condition-based and predictive maintenance

The student is able to:

• ELO CMRAM3 - strategically plan maintenance in industrial plants and energy systems
• ELO CMRAM4 - implement AI and simulation methods for fault detection, diagnostics and prognostics

The student can communicate:

• ELO CMRAM5 - the results of the application of AI and simulation methods to fault detection, diagnostics and prognostics, condition-based and predictive maintenance

INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS

Part 1A Deterministic Safety Analysis (DSA)

• DSA safety concepts (defense-in-depth, multiple barriers, Design Basis Accidents (DBAs) and Beyond DBAs (BDBAs)) and the technological implications on the nuclear power plants design
• Strengths and weaknesses of DSA

Part 2A Probabilistic Safety Analysis (PSA)

• Level 1,2,3 PSA: methods and tools for modeling of nuclear accidents
• Strengths and weaknesses of PSA

Part 3A Integrated Deterministic and Probabilistic Safety Analysis (IDPSA)

• Dynamic accidental scenarios generation:
• Multi-Valued Logic (MVL) for modeling the accidental scenarios
• Continuous Event Tree (CET) and Continuous Cell-to-Cell Mapping Technique (CCCMT)
• Dynamic Event Tree (DET) and Monte Carlo Dynamic Event Tree (MCDET)
• Dynamic accidental scenarios modeling and simulation:
• Uncertainty and sensitivity analysis of Thermal-Hydraulics codes (Code Scaling, Applicability, and Uncertainty (CSAU), Automated Statistical Treatment of Uncertainty Method (ASTRUM)), global variance-based and distribution-based measures (flipped))
• Risk-informed safety margin quantification (Order Statistics (OS))
• Dynamic accidental scenarios post-processing:
• Artificial Intelligence methods (Support Vector Machines (SVMs), Artificial Neural Networks (ANNs)) and advanced simulation methods (Kriging, Advanced Monte Carlo methods) for system failure domain identification
• Artificial Intelligence methods (evolutionary algorithms and clustering algorithms) for prime implicants and near misses scenarios identification

COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE

Part 1B: Introduction

• Maintenance approaches (Unscheduled (corrective) vs scheduled (Planned, On Condition, Predictive and Prescriptive)) and strategic planning (Reliability centered maintenance and risk-based maintenance)

Part 2B: CORRECTIVE AND SCHEDULED MAINTENANCE

• Scheduling the optimal maintenance time by AI (Multi Objective Genetic Algorithms)

Part 3B: ON CONDITION AND PREDICTIVE MAINTENANCE

• Fault detection by Principal Component Analysis and application to fault detection in NPPs
• Fault Diagnostics by AI (Fuzzy C Means, K Nearest Neighbor Clustering Methods; Ensemble Systems) and application to fault diagnostics in NPPs
• Fault Prognostics by experience based methods (Weibull distribution), statistical methods (AutoRegressive Moving Avarage (ARMA), Hidden Markov models, Proportional Hazard Models (PHM)), advanced simulation methods (kalman and particle filtering (flipped)), AI methods (Ensemble System)
• Development and/or application of algorithms for fault detection, diagnostics and prognostics

The courses INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS and COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE are self-consistent (the one is not needed for the other, and viceversa).

The evaluation consists of a final oral exam and of a project work.

The oral exam (70% of the final grade) is aimed at verifying the student:

• ability to expose and discuss the mathematical formalism of the AI and advanced simulation methods presented during the course (ELO IDPSA1-2; ELO MCRAM 2)
• understanding of the need (and opportunity) tackled (and offered) by AI and advanced simulation method for the reliability, safety and maintenance of industrial plants and energy systems (ELO IDPSA 3; ELO MCRAM 1-2)
• ability to communicate knowledge clearly and unambiguously (ELO IDPSA 6; ELO MCRAM 5)

The project work (30% of the final grade), done in groups of 2 students, is aimed at verifying the student ability to develop and apply the AI and advanced simulation methods for the sensitivity analysis of nuclear accidents models and codes, and the dynamic accidental scenarios analysis (course INTEGRATED DETERMINISTIC AND PROBABILISTIC SAFETY ANALYSIS OF NUCLEAR POWER PLANTS), and/or the fault detection, detection, diagnostics and prognostics (course COMPUTATIONAL METHODS FOR RELIABILITY, AVAILABILITY AND MAINTENANCE) (ELO IDPSA 4-5; ELO CMRAM 3-4). The project work will be summarized by the students in a technical scientific report and presented in a scientific talk, to be given during a plenary presentation session to be held in class. The ability to communicate knowledge clearly and unambiguously will be evaluated (ELO IDPSA 6; ELO CMRAM 5)