Fuzzy Cognitive Maps (FCMs) are computational models that combine fuzzy logic and cognitive mapping to represent complex systems and their interrelationships. This thesis provides an overview of FCMs, their evolution, theoretical foundations, and various approaches for constructing and analysing them. FCMs offer a powerful framework for modelling complex systems and decision-making processes, integrating human expertise and computational techniques to navigate multifaceted challenges. The thesis comprises three case studies that showcase the versatility and efficacy of FCMs in addressing real-world problems. The first case study focuses on public project management, where FCMs are used to analyse administrative processes in a public university, with the purpose of enhancing public sector efficiency and decision-making. In the second case study, the focus shifts to the oil&gas sector, where FCMs are combined with machine learning techniques, specifically Gray Wolf Optimisation (GWO), to model and simulate the behaviour of vertical tanks in multiphase liquid-gas plants. This innovative approach offers a novel methodology for accurate estimation and predictive modelling, leveraging computational algorithms to navigate the complexities of industrial processes. The third case study explores the critical role of anomaly detection systems in oil&gas plants, employing FCMs and GWO to identify and categorise anomalies in real-time. By integrating advanced algorithms, this study demonstrates the potential of FCMs in enhancing operational resilience and decision support, while minimising false alarms and optimising energy management practices. Collectively, these case studies contribute to advancing the understanding and application of FCMs in diverse domains, showcasing their effectiveness in addressing contemporary complex challenges and informing strategic decision-making processes.
Le Mappe Cognitive Fuzzy (FCM) sono modelli computazionali che combinano la logica fuzzy e la mappatura cognitiva per rappresentare sistemi complessi e le loro interrelazioni. Questa tesi fornisce una panoramica delle FCM, della loro evoluzione, delle fondamenta teoriche e dei vari approcci per costruirle e analizzarle. Le FCM offrono un potente quadro per modellare sistemi complessi e processi decisionali, integrando l'esperienza umana e le tecniche computazionali per affrontare sfide audaci. La tesi comprende tre casi studio che mostrano la versatilità e l'efficacia delle FCM nel risolvere problemi reali. Il primo studio di caso si concentra sulla gestione di progetti pubblici, dove le FCM vengono utilizzate per analizzare i processi amministrativi in un’università pubblica, con l’obiettivo di migliorarne l'efficienza e le procedure decisionali. Nel secondo caso studio, l'attenzione si sposta sul settore oil&gas, dove le FCM sono combinate con tecniche di apprendimento automatico, in particolare con la Gray Wolf Optimisation (GWO), per modellare e simulare il comportamento dei serbatoi verticali in impianti liquido-gas multifase. Questo approccio offre una metodologia nuova per stime accurate e modellazione predittiva, sfruttando algoritmi computazionali per navigare nelle complessità dei processi industriali. Il terzo caso studio esplora il ruolo dei sistemi di rilevamento delle anomalie negli impianti oil&gas, impiegando FCM e GWO per identificare e categorizzare anomalie in tempo reale. Integrando algoritmi avanzati, questo studio dimostra il potenziale delle FCM nel migliorare la resilienza operativa e il supporto decisionale, riducendo al minimo i falsi allarmi e ottimizzando le pratiche di gestione energetica. Complessivamente, questi casi studio contribuiscono ad avanzare la comprensione e l'applicazione delle FCM in diversi ambiti, mostrando la loro efficacia nell’affrontare sfide contemporanee complesse e nell’informare i processi decisionali strategici.
Exploring Fuzzy Cognitive Maps: A Dual Approach Integrating Qualitative Expertise and Quantitative Methods / Carbonari, Sara. - (2024 Jun 17).
Exploring Fuzzy Cognitive Maps: A Dual Approach Integrating Qualitative Expertise and Quantitative Methods
CARBONARI, SARA
2024-06-17
Abstract
Fuzzy Cognitive Maps (FCMs) are computational models that combine fuzzy logic and cognitive mapping to represent complex systems and their interrelationships. This thesis provides an overview of FCMs, their evolution, theoretical foundations, and various approaches for constructing and analysing them. FCMs offer a powerful framework for modelling complex systems and decision-making processes, integrating human expertise and computational techniques to navigate multifaceted challenges. The thesis comprises three case studies that showcase the versatility and efficacy of FCMs in addressing real-world problems. The first case study focuses on public project management, where FCMs are used to analyse administrative processes in a public university, with the purpose of enhancing public sector efficiency and decision-making. In the second case study, the focus shifts to the oil&gas sector, where FCMs are combined with machine learning techniques, specifically Gray Wolf Optimisation (GWO), to model and simulate the behaviour of vertical tanks in multiphase liquid-gas plants. This innovative approach offers a novel methodology for accurate estimation and predictive modelling, leveraging computational algorithms to navigate the complexities of industrial processes. The third case study explores the critical role of anomaly detection systems in oil&gas plants, employing FCMs and GWO to identify and categorise anomalies in real-time. By integrating advanced algorithms, this study demonstrates the potential of FCMs in enhancing operational resilience and decision support, while minimising false alarms and optimising energy management practices. Collectively, these case studies contribute to advancing the understanding and application of FCMs in diverse domains, showcasing their effectiveness in addressing contemporary complex challenges and informing strategic decision-making processes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.