Tesi di Dottorato
Permanent URI for this communityTesi di Dottorato
Browse
5 results
Search Results
Item Model-based and simulation-driven methods for the reliability and safety analysis of systems(2013-11-28) Tundis, Andrea; Greco, Sergio; Garro, AlfredoIn several industrial domains such as automotive, railway, avionics, satellite, health care and energy, a great variety of systems are currently designed and developed by organizing and integrating existing components (which in turn can be regarded as systems),that pool their resources and capabilities together to create a new system which is able to o er more functionalities and performances than those o ered by the simple sum of its components. Typically, the design and management of such systems, whose properties cannot be immediately de ned, derived and easily analyzed starting from the properties of their parts when they are considered in stand-alone, require to identify and face with some important research issues. In particular, the integration of system components is a challenging task whose criticality rises as the heterogeneity and complexity of the components increase. Thus, suitable engineering methods, tools and techniques need to be exploited to prevent and manage the risks arising from the integration of system components and, mainly, to avoid their occurrence in the advanced phases of the system development process which may result in a signi cant increase in the entire project costs. To overcome these issues the adoption of the Systems Engineering approach represents a viable solution as it provides a wide set of methods and practices which allow the de nition of the system architecture and behavior at di erent abstraction level in terms of its components and their interactions. Moreover, systems requirements are constantly traced during the di erent system development phases so to clearly specify how a system component concurs to the ful llment of the requirements. However, in the Systems Engineering eld, even though great attention has been devoted to functional requirements analysis and traceability, there is still a lack of methods which speci cally address these issues for non-functional requirements. As a consequence, the analysis concerning if and how non-functional requirements are met by the system under development is not typically executed contextually to the design of the system but still postponed to the last stages of the development process with a high risk of having to revise even basic design choices and with a consequent increase in both completion tim and development costs. Among all system requirements, Reliability and Safety are important non-functional requirements. Especially for mission-critical systems, there is a strong demand for new and more powerful analysis tools and techniques able not only to verify the reliability indices and safety of a system but also to exibly evaluate the system performances and compare di erent design choices. In this context, the research aimed to promote the use of exible methods for the analysis of non-functional requirements by focusing on the de nition of: (i) model-based method for system reliability analysis centered on popular SysML/UML-based languages for systems modeling and on de-facto standard platforms for the simulation of multi-domain dynamic and embedded systems (Mathworks Simulink); (ii) a methodological process for supporting the safety analysis, along with an approach for performing the Fault Tree Analysis of cyber-physical systems, mainly based on the Modelica language and OpenModelica simulation environment. Furthermore, in order to support the representation of system requirements and thus enable their veri cation and validation during the design stages, a meta-model for modeling requirements of physical systems as well as di erent approaches for extending the Modelica language have been proposed. Moreover, an algorithm, which allows trace and evaluate requirements violation through simulation, has been de ned. Finally, the e ectiveness of the proposed methods and approaches, especially in the modeling and analysis of both the expected and dysfunctional system behavior, is the result of an intensive experimentation in several industrial domains such automotive, avionics and satelliteItem Classification models and algorithms in application of multi-sensor systems to detection and identification of gases(2014-06-04) Khalaf, Walaa; Cocurullo, Giuseppe; Gaudosio, Manlio; Pace, CalogeroThe objective of the thesis is to adopt advanced machine learning tech- niques in the analysis of the output of sensor systems. In particular we have focused on the SVM (Support Vector Machine) approach to classi- ¯cation and regression, and we have tailored such approach for the area of sensor systems of the "electronic nose" type. We designed an Electronic Nose (ENose), containing 8 sensors, 5 of them being gas sensors, and the other 3 being a Temperature, a Humidity, and a Pressure sensor, respectively. Our system (Electronic Nose) has the ability to identify the type of gas, and then to estimate its concentration. To identify the type of gas we used as classi¯cation and regression technique the so called Support Vector Machine (SVM) approach, which is based on statistical learning theory and has been proposed in the broad learning ¯eld. The Kernel methods are applied in the context of SVM, to improve the classi¯cation quality. Classi¯cation means ¯nding the best divider (separator) between two or more di®erent classes without or with minimum number of errors. Many methods for pattern recognition or classi¯cation are based on neural network or other complex mathematical models. In this thesis we describe the hardware equipment which has been designed and implemented. We survey the SVM approach for machine learning and report on our experimentation.Item Metodi di rilevazione ed isolamento guasti per sistemi LPV ed ibridi(2014-03-07) Gagliardi,Gianfranco; Casavola,Alessandro; Palopoli,Luigi; Famularo,DomenicoItem Model predictive control schemes for linear parameter varying systems(2014-03-03) Garone,Emanuele; Talia,Domenico; Casavola,AlessandroThis dissertation presents several contributions inherently the control of con- strained Linearly Parameter Varying (LPV) systems. First the basic analysis and synthesis tools needed to deal with the class of LPV systems are carried out and introduced. Some novel results are given, especially for what regards the use of scheduled control laws and stability conditions for LPV with slow parameter variations. Then we moved on the problem of constrained control. Several new constrained stabilization results are proposed here for the first time and improvements in the procedures to build-up time-variant strategies able to deal with constrained LPV system are given. Moreover a new particular kind of control strategy based on the idea of exploiting the prediction set structure is introduced here for the first time in the LPV framework. It has been pointed out in which way those approaches can be arranged within Model Predictive Control (MPC) schemes to more efficiently deal with constrained LPV systems and two new fast-MPC algorithms for LPV system have been proposed. In this thesis some attention has been given to the analysis of LPV systems with slow parameter variations and some preliminary results are reported. Such a class of systems has many potentials but, due to the ”hidden” nonlinearities it introduces, it is still not well understood and would deserves further careful investigations.Item Nuovi modelli matematici per il dimensionamento degli scambiatori di calore terreno-aria(2014-02-14) Vulcano,Antonio; Cucumo,Mario A.; Luchi,Maria L.