| KEYWORD |
Stochastic Simulation and Machine Learning for the Resilience Analysis of Energy Critical Infrastructures (ECI) in the Presence of Disruptive Events
Parole chiave ADAPTIVE SIMULATION, ARTIFICIAL INTELLIGENCE, CLIMATE CHANGE RESILIENCE, CRITICAL INFRASTRUCTURES PROTECTION, ENERGY CRITICAL INFRASTRUCTURE, MACHINE LEARNING, NATURAL HAZARDS, RESILIENCE, STOCHASTIC SIMULATION
Riferimenti NICOLA PEDRONI
Gruppi di ricerca Nemo
Descrizione Context:
Modern society depends on critical lifelines, like electric power and natural gas, telephone and other communication systems, water, wastewater and transportation systems. Many of these infrastructures are aging and are exposed to different natural hazards (e.g., earthquakes, tsunamis, hurricanes and floods), whose intensity and severity has been increasing in the past years, possibly due to climate change. Resilience has become a relevant attribute to consider in the design for the management of any infrastructure, for ensuring the capability of withstanding disruptive events and recovering performance.
Scenario simulation plays a key role for analyzing resilience and informing investments on prevention, mitigation and recovery. However, it becomes computationally demanding when exploring the response of the complex Critical Infrastructures (CI) under uncertain disruptive events. In this view, Artificial Intelligence (AI), Machine Learning (ML) and stochastic simulation can be combined for simulating efficiently the response of an infrastructure exposed to a disruptive event, under uncertain operational and environmental conditions.
Activities:
Within this general framework, we will focus on the Energy Critical Infrastructures (ECI) for their pivotal role in supporting other infrastructures and society as a whole, in a scenario of energy transition. The main activities of the MSc student will be the following:
- carry out a synthetic literature review about the problem of resilience (analysis and quantification) in Energy Critical Infrastructures (ECIs);
- develop and integrate stochastic simulation and AI/ML approaches for the intelligent, adaptive exploration and identification of disruption and recovery scenarios, associated to extreme natural events possibly affecting the ECI (in particular, an earthquake and/or a tsunami);
- apply the developed method to an Integrated Energy System (combining a Gas Turbine Plant, a Nuclear Power Plant, and a Wind Farm), whose model can be made available to the student;
- based on the information collected from the AI/ML-based stochastic exploration, perform a preliminary selection of optimal recovery strategies to maximize post-disruption system resilience.
Vedi anche resilience_eci_artificialintelligence_stochasticsimulation - 2.pdf
Note Duration of the thesis
Around 6-8 months (anyway, this aspect can/should be discussed)
Scadenza validita proposta 31/12/2025
PROPONI LA TUA CANDIDATURA