Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)

Permanent URI for this collection

https://hdl.handle.net/10889/21

Browse

Now showing 1 - 20 of 370

Open Access
A cognition-centered personalization framework for cultural-heritage applications
Ράπτης, Γεώργιος; Raptis, George
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
Cultural-heritage applications provide the visitors with high volume of cultural information and therefore there is a need for personalizing the visit experience to help visitors understand and perceive the provided cultural content according to the characteristics of their cognitive style. However, the current research and design attempts do not consider the diversity of the visitors, in terms of their cognitive styles, as an important design and evaluation factor. Cognitive styles influence the way people seek, collect, organize, process, and recall information and are expected to have an impact on the cultural-heritage domain, which is enriched in cultural information. The results presented in the PhD thesis revealed that visitors with different cognitive styles behave differently when interacting with cultural-heritage applications. Theses differences influence the visit experience, in terms of content understanding and immersion. The current cultural-heritage applications do not consider the diversity of the visitors in terms of cognitive styles, and thus, they unintentionally led to experience imbalances between visitors with different cognitive styles. The PhD thesis presents CogniCHeF, which is a framework that aims to help designers to deliver personalized cultural-heritage applications tailored to the visitors’ cognitive styles, aiming to offer them an enhanced visit experience. Moreover, the results presented in the PhD thesis revealed that the cognitive styles can be elicited automatically with high accuracy during the first stages of interaction between the visitors and the cultural-heritage applications through the use f eye-tracking data. Furthermore, the results presented in the PhD thesis revealed that CogniCHeF can be efficiently used by cultural-heritage designers to implement personalized applications tailored to the visitors’ cognitive styles. Finally, the results presented in the PhD thesis revealed that cognition-centered personalization improved the overall visit experience.
Open Access
A new framework for structuring and deploying advanced personalized and ubiquitous healthcare services
(2014-08-03) Φέγγου, Μαρία-Άννα; Λυμπερόπουλος, Δημήτριος; Λυμπερόπουλος, Δημήτριος; Κωτσόπουλος, Σταύρος; Κουκιάς, Μιχαήλ; Δενάζης, Σπύρος; Σγάρμπας, Κυριάκος; Αναστασόπουλος, Γιώργος; Παναγιωτάκης, Γιώργος; Fengou, Maria-Anna
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This dissertation focuses on the design of personalized and ubiquitous healthcare (PUH) services in the medical sector. It is proposed a new framework for structuring and deploying advanced PUH services in the right place, time and manner. According to this PUH framework, diverse types of entities (subjects and objects) are involved during the execution of each service. Each type of subject (e.g. patient, caregiver (doctor or nurse), volunteer, patient’s friend or relative, etc) or object (medical unit, medical files manager, etc) performs predefined actions according to the patient’s current health status and an agreed plan of actions. In the provision of PUH services, the permissible actions of each entity (or a group of entities) are determined by the “entity’s profile”, which is created in advance for this purpose. This work considers also the “group profile” that determines the behavior and the attributes of a specific group of entities. The PUH framework provides also mechanisms for the management of the patient’s context (environmental, living, etc) and the content of subject’s profile. Different profiling mechanisms are activated according to the subject’s type. Structurally, the proposed framework is an expansion of the framework that is utilized by the Next Generation Networks (NGN) (known as OSA/Parlay architecture) for the reliable provision of multipoint – multimedia communication services. The PUH framework distinguishes four classes of mechanisms: a) the conventional ETSI/Parlay mechanisms handling the communication among the entities that are involved during the execution of a PUH service, b) the mechanisms handling the acquisition of bio and contextual data by sensor networks deployed around the patient (e.g. Body Area Network), c) the mechanisms for the management of profiles data (e.g. availability, preferences, capabilities, care giving role, activities) and d) the security mechanisms. For the representation of these classes of mechanisms, an appropriate ontology has been created. Based on the above PUH framework, an entire profile management system is proposed. This system uses data from patient/caregivers profiles (or group profiles) in order to deploy a PUH service that will be able to support multi-party group-working schemes with well defined behaviors of all involved entities, devices and services. The technology of smart cards has selected as the appropriate technology for creating, accessing and enriching the subjects profiles. A prototype of the proposed profile management system is implemented using the cloud computing technology (Windows Azure platform). We have considered that parts of a profile are located in different service providers. The efficiency of the developed profile management system is evaluated in terms of time response in order to select the appropriate healthcare provider by simulating system's response to real world scenarios.
Open Access
Acoustic energy harvesting with emphasis on low frequencies
Παπαδάκος, Χαράλαμπος; Μουρτζόπουλος, Ιωάννης; Φακωτάκης, Νίκος; Καλύβας, Γρηγόριος; Πολύζος, Δημοσθένης; Μενούνου, Πηνελόπη; Πυργιώτη, Ελευθερία; Μπίρμπας, Μιχαήλ; Papadakos, Charalampos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This work presents novel findings and results in the field of acoustic energy harvesting with emphasis on low frequencies. At the beginning of this work, ambient acoustic power density is estimated and a comparison between the potential of sound and that of other ambient energy sources exploited for energy harvesting is performed. The interest in AEH is justified and the corresponding state of the art is presented. Prior the analysis, the motivation of this work is described in detail and the main objectives are set. The analysis considers the piezoelectric transducer to be the transduction element for this study. Initially, it provides a lumped element model (LEM) for evaluating the AEH potential of such transducers directly coupled to the sound field and utilizing such a framework, it demonstrates the inefficiency of such approach for AEH of practical interest. Subsequently, a novel method and device is proposed for improving the coupling of such transducers and the sound field and enhancing the potential of AEH. For this case of indirect coupling, the analysis provides a modified LEM compared to the direct coupling case. At a next level, a novel biomimetic device that mimics the function and the structure of the human middle ear is proposed for implementing such indirect coupling of piezoelectric transducers and the sound filed for AEH. Such device can potentially ensure sufficient electric power for driving low power applications in the range nW - mW. Next, in order to provide an AEH proof - of - concept case, this study focuses on loudspeaker enclosures since the sound pressure level (SPL) therein can be significantly high during operation, as it is shown via tests performed for realistic 3D virtual prototype models, which are validated also via measurements performed inside real - life loudspeaker systems. Specific examples of 3D virtual prototype models of loudspeaker enclosures are combined with the proposed biomimetic device and a piezoelectric transducer and are tested for evaluating the in - Box AEH potential. The test results indicate that the proposed biomimetic device may operate as a harvester and the same time as a passive absorber. These test results further show that the output voltage of the proposed AEH can be adequate for charging a rechargeable battery or for driving an electronic load, while the electric power delivered to a load resistance may be up to units of mW. These results presented in this work are discussed in detail and the framework for future work and applications is presented.
Open Access
Adaptive polarization mode dispersion equalizers for coherent optical communications systems
(2010-11-01T07:32:17Z) Μαντζούκης, Νικόλαος; Ρούδας, Ιωάννης; Ρούδας, Ιωάννης; Τζες, Αντώνιος; Μπερμπερίδης, Κωνσταντίνος; Γκίκας, Δημήτριος; Μπίρμπας, Αλέξιος; Καμαλάκης, Θωμάς; Βλάχος, Κυριάκος; Mantzoukis, Nikolaos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
Polarization mode dispersion (PMD) arises as a result of the birefringence in optical fibers, due to inherent asymmetries and deformities from external stresses. The spectral components of the input optical pulse propagate with different group velocities. Consequently, pulse duration increases leading to intersymbol interference between consequent symbols, leading to performance reduction of the coherent systems. In order to compensate for the PMD, we use adaptive linear PMD equalizers. Due to the dynamic and random nature of PMD, it is crucial for a system designer to efficiently simulate the PMD-induced outage probabilities of 10-5. Because of this stringent requirement, it is computationally costly to use the conventional Monte Carlo methods. To overcome this hurdle, Importance Sampling methods, such as the multicanonical Monte Carlo method have been applied in the past in order to efficiently reduce the simulation time required to estimate the statistics of these rare events. The multicanonical Monte Carlo method does not require any prior knowledge of which rare events contribute significantly to the PMD-induced outages. In essence, multicanonical Monte Carlo simulations adaptively bias the input random variables with a priori unknown weights. The PMD emulation model consists of a concatenation of birefringent sections, simulated based on MMC. The objective of this dissertation is to apply, for the first time, the multicanonical Monte Carlo method to accurately and efficiently evaluate the performance of adaptive, blind, feed-forward PMD equalizers employed in coherent polarization division multiplexed (PDM) quadrature phase-shift keying (QPSK) systems in all order PMD emulation model. In the exclusive presence of PMD, we demonstrated that the half-symbol-period-spaced adaptive electronic equalizers, based on the constant modulus algorithm (CMA) equalizers perform slightly better than the decision directed least mean square (DD-LMS) counterparts at links with larger PMD values, whereas the opposite holds true for the low PMD regime. Due to their distinguishable performance in different regimes of the PMD, they provided an even better performance when running DD-LMS after a first round of CMA-based equalization than using either one of the equalization algorithms stand alone. Finally, the joint presence of PMD and intermediate frequency offset or PMD and random differential phase carrier shifts slightly worsened the performance of the coherent PDM QPSK systems, independently of the equalizer. Although these random differential carrier phase shifts are typically omitted in similar PMD studies in intensity modulated/direct detection (IM/DD) systems, they should be taken into account in due to the phase sensitivity of the PDM QPSK coherent systems.
Open Access
An architecture to support design of context-aware mobile applications
(2011-06-16T06:52:50Z) Stoica, George Adrian; Αβούρης, Νικόλαος; Αβούρης, Νικόλαος; Σερπάνος, Δημήτριος; Φακωτάκης, Νικόλαος; Milrad, Marcelo; Κόμης, Βασίλειος; Παπαθεοδώρου, Χρίστος; Λέπουρας, Γεώργιος
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This thesis deals with the problem of interaction with context-aware mobile applications in digitally augmented physical spaces. The term "mobile application" is used as an umbrella term for any piece of software that can run on a mobile device in order to support a human task. “Context-aware” refers to the ability of such applications to use information about the environment or the user to adapt their behaviour accordingly. A "physical hyperlink" is a mechanism that associates a physical object or point in space with digital information/service(s). The term “digitally augmented physical space” refers here to a physical space where digital information relevant to points or artefacts can be accessed (e.g. by means of physical hyperlinks). In the last two decades mobile devices and applications have developed tremendously. With them, on one hand new ways appeared to support human tasks and on the other hand new research challenges emerged. A key factor is the continuous change of environment and context of use. This is very different to a classic desktop computer application, in which the external environment is hardly changing, as the user and the computer do not change location during interaction. In the world of mobile computing the environment and the context of use are changing, at a faster rate. One of the most widely spread mobile devices, the mobile phone, evolved into a powerful mobile computing device. These advanced devices are now ubiquitously spread and already billions of people carry and use them daily. New services and uses are emerging with every step of evolution of this technology. However these rapidly emerging and evolving technologies and services result in a generally heterogeneous environment that affects user experience. The advances in mobile technology permitted devices to be equipped with sensors and hardware that allow on one hand the acquisition of information from the environment and on the other hand interaction with objects like physical hyperlinks, that permit direct access to information related to physical artefacts. In the frame of this thesis a conceptual framework (MobiAct framework) and a service oriented architecture (MobiAct architecture) have been defined. The MobiAct framework deals with the interaction with context-aware mobile applications in a digitally augmented physical space. The MobiAct framework has as a goal to identify the main concepts that come into play while interacting with context-aware applications in digitally augmented physical spaces. MobiAct architecture is an innovative solution for delivering context-aware mobile services in digitally augmented physical spaces. The MobiAct architecture has MobiAct framework as the underlying base. The architecture aims at enabling a uniform and consistent user experience across spaces and contexts and at facilitating the design and development of context-aware mobile services by providing ready functionality for common tasks. Finally a set of software architectural patterns relevant to designing mobile context-aware applications have been identified. These patterns provide a set of building blocks to facilitate the design and development of similar systems. The road towards designing MobiAct conceptual architecture and framework for interacting with context-aware mobile applications included the development and evaluation of several prototypes mainly targeted at public and semi-public environments. These prototypes contributed in a great degree towards exploring and studying context-aware mobile applications and interaction with digitally augmented physical spaces using physical hyperlinks and subsequently defining the requirements for the architecture. To avoid binding the architecture to capabilities of a certain platform, the prototypes involved different devices and platforms. The prototypes and applications developed included educational games and a guide for a museum, an application to support users of a library and a shopping assistant aimed at a supermarket environment. The findings of these efforts contributed in the definition of the MobiAct architecture that has been evaluated through a software architecture analysis method based on scenarios and discussions involving representative stakeholders as well as an evaluation team
Open Access
Biomechanical simulation of virtual physiological humans : modeling of musculoskeletal kinematic and dynamic redundancy using coordinate projection methods
Stanev, Dimitar; Μουστάκας, Κωνσταντίνος; Δερματάς, Ευάγγελος; Κουτσογιάννης, Κωνσταντίνος; Τζες, Αντώνιος; Μπαλτζόπουλος, Βασίλειος; Τζοβάρας, Δημήτρης; Δεληγιάννη, Δέσποινα; Στάνεβ, Ντιμιτάρ
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The principles of musculoskeletal model and simulation have received much attention over the last decades, enabling the prediction of surgical treatments related to different motion limiting disorders. However, their application in clinical practice is still limited partly because the experimental equipment used for measuring the kinematics and kinetics required for the analysis is expensive and obtrusive. The assessment of the internal state (e.g., muscle forces and joint reaction loads) from those measurements does not lead to a unique solution, due to the inherit redundancy of the musculoskeletal system. More specifically, there are more degrees of freedom than those required to perform certain tasks and each degree of freedom is actuated by multiple muscles, leading to infinite combinations of muscle forces that satisfy the movement. Unfortunately, this raises the following questions: which is the true solution employed by the central nervous system and whether the choice of a particular solution can lead to misinterpretation of results? Coordinate projection methods and their extension for musculoskeletal modeling and simulation are the topic of this thesis. These methods transform the equations of motion into a space of low- or high-dimensionality according to the projection operator. Five different subspaces are studied, namely task, joint, muscle, constraint and null space as well as their relationship. Task space projection simplifies the motion planning problem and the process of synthesizing virtual simulations. This is of great importance, since simulations can be arranged effortlessly and intuitively. Joint space representation is the de facto standard for formulating the underlying equations of motion and dynamics simulation methods. Muscle space projection provides a convenient representation for interfacing musculoskeletal and segmental level models and forms a basis for practical control applications. Constraint projection serves to incorporate the kinematic constraints into the inverse dynamics model. Null space projection can be used to model the redundancy effects of the musculoskeletal system consistently and identify the feasible solution that satisfy the movement and physiological muscle constraints. The redundant nature of the musculoskeletal system introduces variability/uncertainty in simulated quantities leading to misinterpretation of the results if ignored. Therefore, this groundwork provides the appropriate formalization to successfully address these issues, facilitating the application of broader types of studies in the realm of motor coordination.
Open Access
Brain signal and image analysis using machine learning methods
Κανάς, Βασίλειος; Σγάρμπας, Κυριάκος; Σγάρμπας, Κυριάκος; Δερματάς, Ευάγγελος; Μπεζεριάνος, Αναστάσιος; Μεγαλοοικονόμου, Βασίλειος; Μπόρας, Ιωσήφ; Μουστάκας, Κωνσταντίνος; Φακωτάκης, Νικόλαος; Kanas, Vasileios
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The work presented in this Ph.D manuscript focuses on implementing and designing machine learning for brain signal analysis. In the first part of the manuscript, dedicated to one-dimensional brain signal, we study electrocorticography (ECoG) signal processing for voice activity detection and syllable classification in order to design an interpretable and more efficient brain computer interface system for speech rehabilitation. The second part of this Ph.D dissertation is dedicated to two-dimensional biomedical signal analysis (image analysis). More specifically, we per-formed analysis of magnetic resonance medical images for brain tumor segmentation and grade classification. Finally, the last part of the thesis is based on mathematical modeling of biological neural networks. We aimed to study the microscopic dynamics of the brain neuronal networks through synchronization phenomena.
Open Access
Control and nonlinear stability analysis of the electromechanical system of electric vehicles
Μακρυγιώργου, Δήμητρα; Makrygiorgou, Dimitra
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The present PhD dissertation is addressed in the research field of control and analysis of motion systems in vehicular industry and more precisely in providing a systematic procedure for the control design applied on electrified vehicles with respect to the system stability. In particular, taking into account the nonlinear representations of all components involved in an electrified vehicle, namely the energy storage, both electric and internal combustion motors, and power converters, an accurate model is developed that sets the basis for control design and analysis purposes. System performance is examined in a first place via the proposal of a novel open-loop stability analysis. The latter is conducted by applying Lyapunov based nonlinear techniques that actually prove system convergence to its desired equilibrium for any feasible bounded control input signals. Those theoretical results set the cornerstone towards adopting intelligent control techniques, fuzzy and neurofuzzy, due to the bounded control signals they produce and their efficiency in copying with system nonlinearities and uncertainties. In a similar manner, a combination of sliding mode and field oriented techniques are applied and provide an improved system dynamic behavior endowing it with characteristics of robustness with the system stability again to be proven based on the novel open-loop analysis. Towards extending the aforementioned effort, a general systematic procedure for control design based on simple cascaded PI controllers that can ensure stability for a wide category of Euler-Lagrange systems is developed. In this framework, asymptotic stability is proven for the electrified vehicle system by keeping in mind its desired equilibrium and applying proper cascaded PI controllers with their gains to be tuned based on the adopted systematic analysis. In addition, a vehicle to grid system is also considered and analysed based on the proposed method and its accurate dynamic representation. Specifically, in a first stage the mathematical representation of a vehicle to grid system combined with multiple connected vehicles is deployed and its stability is proven by applying trivial PI controllers and adopting the proposed open-loop analysis. Afterwards, a unified vehicle to grid - plug-in electric vehicle system is developed and controlled based on the proposed systematic passivity based procedure that ensures global asymptotic stability of the entire system around its desired equilibrium. In all cases, the extracted theoretical results are verified by conducting extensive simulations that indicate a satisfactory system response with smooth transients and a successful driving of the entire system to its desired equilibrium.
Open Access
Control of cooperative unmanned aerial vehicles
(2011-10-06T08:47:37Z) Αλέξης, Κώστας; Τζες, Αντώνιος; Ευσταθίου, Κωνσταντίνος; Βαλαβάνης, Κίμων; Κουμπιάς, Σταύρος; Ασπράγκαθος, Νικόλαος; Δερματάς, Ευάγγελος; Μάνεσης, Σταμάτιος; Alexis, Kostas
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This thesis addresses the problems of design and control of small cooperative unmanned autonomous quadrotor aerial vehicles. A new approach is proposed, for the modeling of the system’s dynamics using linearized Piecewise AffineModels. The Piecewise Affine dynamic–models cover a large part of the quadrotor’s flight envelope while also taking into account the additive effects of environmental disturbances. The effects of aerodynamic forces and moments were also examined. A small quadrotor is designed and developed that emphasizes in the areas of increased on–board computational capabilities, state estimation and modular connectivity. Based on the translational and rotational system’s dynamics: a) a switching model predictive controller, b) an explicitly solved constrained finite time optimal control strategy, and c) a cascade control scheme comprised of classical Proportional Integral Derivative control scheme augmented with angular acceleration feedback, were designed and experimentally tested in order to achieve trajectory tracking under the presence of wind–gusts. The efficiency of the proposed control methods was verified through extended experimental studies. The final quadrotor design utilizes a powerful control unit, a sensor system that provides state estimation based on inertial sensors, ultrasound sonars, GPS and vision chips, and an efficient actuating system. The research effort extended in the field of unmanned aerial vehicles cooperation. Cooperation strategies were proposed in order to address the problems of: a) Forest Fire Monitoring and b) Unknown Area Exploration and Target Acquisition. The Forest FireMonitoring algorithm is formulated based on consensus systems theory formulated as a spatiotemporal rendezvous problem in between the quadrotors. The Area Exploration and Target Acquisition algorithm is formulated based on market–based approaches.
Open Access
Control of networked autonomous robotic vehicles
Γιαννουσάκης, Κωνσταντίνος; Giannousakis, Konstantinos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This dissertation addresses the control of networked autonomous mobile robots and its experimental evaluation. The mobile robots are deployed either as a swarm or as standalone agents in a convex region, or a general non-convex region with obstacles. The topics that we examine are trajectory planning, motion control, obstacle avoidance, area coverage and homing. The fundamental contributions of this dissertation are as follows. Firstly, an automated piecewise optimal trajectory generation method for the mobile robots is developed. Secondly, an obstacleaware motion controller for the mobile robots is developed. Thirdly, a visibility based cell assignment method is proposed and used for the derivation of optimal distributed control schemes for the tasks of area coverage and homing in non-convex regions. Apart from that, each task is evaluated on a network of real robotic vehicles, where the robots are controlled in real time by properly exploiting computational geometry tools. The experiments are conducted using mobile robots with differential drive, with the support of a self-developed external localization system. This camera based localization system is implemented in an augmented reality framework and is included in the network of the experimental setup as the main way of retrieving the position and orientation of the robots in the region of interest. The resulting position and orientation of each robot is combined with its encoder measurements in order to obtain its full state. Additionally, since the effective control of the robots is an absolute necessity, the kinodynamic model of the differential drive robot along with its kinematic constraints are considered throughout the dissertation. The most straightforward method of controlling a real robot with respect to its model is through trajectory planning. After defining the trajectory function and determining the required properties, we propose an automated time-optimal trajectory generation method, considering a given set of points to be traversed. Based on the initial pose and velocity of the robot, our method connects the points of interest in sequence by cubic Bezier segments. Using non-linear constrained optimization, these segments are time-optimized under the velocity and acceleration constraints of the robot. The optimization problem is simplified through root-finding and evaluation of polynomials and results in an applicable, smooth trajectory, that exploits the dynamic capabilities of the robot as much as possible. Even though trajectories can easily incorporate the robot dynamics, they are not an eﬀicient way of controlling the robots of a swarm, where the motion of the robots is determined by their interactions. In most tasks of swarm robotics, the high level controller of each robot provides a desirable movement direction, thus a motion controller is needed to translate this direction into the actual control input of the robot. As a first step, we designed a region unaware motion controller that drives the robot as fast as possible towards a desired direction, considering the dynamic constraints, the sampling interval and the maximum traverse distance. This controller is further extended into a region aware controller that, apart from the dynamics and constraints, considers the robot dimensions as well, in order to avoid the obstacles in the visibility field of the robot. The performance of these controllers is evaluated in experimental studies, through the task of position control, in a convex and in a non-convex region, where the movement direction is determined by the shortest path connecting the robot to its target. Finally, relying on the aforementioned motion controllers, we examined the distributed control of multi-robot configurations. After experimental evaluation of certain already developed area coverage schemes for convex regions, we turned our focus towards the distributed control in non-convex regions with obstacles. Initially, a method of assigning parts of the non-convex region to the robots was developed, where the parts in the vicinity of each robot are determined in terms of the visibility notion. Based on that, optimal spatially distributed coordination algorithms were successfully derived for the homing problem, where the robots should reach some known locations, as well as the area coverage problem, after properly expressing each of them by an aggregate objective function.
Open Access
Design, control and autonomous navigation of tilt-rotor unmanned aerial vehicles
Παπαχρήστος, Χρήστος; Τζες, Αντώνιος; Τζες, Αντώνιος; Βαλαβάνης, Κίμων; Γρουμπός, Πέτρος; Δερματάς, Ευάγγελος; Κούσουλας, Νικόλαος; Κυριακόπουλος, Κωνσταντίνος; Μάνεσης, Σταμάτιος; Papachristos, Christos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This Dissertation addresses the design and development of small-scale Unmanned Aerial Vehicles of the TiltRotor class, alongside their autonomous navigation requirements, including the fully-onboard state estimation, high-efficiency flight control, and advanced environment perception. Starting with an educated Computer Assisted Design-based methodology, a mechanically robust, customizable, and repeatable vehicle build is achieved, relying on high-quality Commercially Available Off-The Shelf equipment –sensors, actuators, structural components–, optionally aided by Rapid Prototyping technology. A high-fidelity modeling process is conducted, incorporating the rigid-body dynamics, aerodynamics, and the actuation subsystem dynamics, exploiting fist- principle approaches, Frequency Domain System Identification, as well as computational tools. Considering the most significant phenomena captured in this process, a more simplified PieceWise Affine system model representation is developed for control purposes –which however incorporates complexities such as flight (state) envelope-associated aerodynamics, the differentiated effects of the direct thrust-vectoring (rotor-tilting) and the underactuated (body-pitching) actuation authorities, as well as their interferences through rigid-body coupling–. Despite the switching system dynamics, and –as thoroughly elaborated– their reliance on constrained manipulated variables, to maintain a meaningful control- oriented representation, the real-time optimal flight control of the TiltRotor vehicle is achieved relying on a Receding Horizon methodology, and more specifically an explicit Model Predictive Control framework. This synthesis guarantees global stability of the switching dynamics, observance of state and control input constraints, response optimality, as well as efficient execution on low computational power modules due to its explicit representation. Accompanied by a proper Low-and-Mid-Level Control synthesis, this scheme provides exceptional flight handling qualities to the aerial vehicle, particularly in the areas of aggressive maneuvering and high-accuracy trajectory tracking. Moreover, the utility of TiltRotor vehicles in the field of aerial robotic forceful physical interaction is researched. Exploiting the previously noted properties of the PieceWise Affine systems Model Predictive Control strategy, the guaranteed-stability Free-Flight to Physical-Interaction switching of the system is achieved, effectively bringing the aerial vehicle into safe, controlled physical contact with the surface of structures in the environment. More importantly, employing rotor-tilting actuation –collectively and differentially– significant forces and moments can be applied onto the environment, while via the standard underactuated authority the vehicle maintains a stable hovering-attitude pose, where the system’s disturbance rejection properties are maximized. Overall, the complete control framework enables coming into physical contact with environment structures, and manipulating the enacted forces and moments. Exploit- ing such capabilities the TiltRotor is used to achieve the execution of physically- demanding work-tasks (surface-grinding) and the manipulation of realistically- sized objects (of twice its own mass) via pushing. Additionally, the fully-onboard state estimation problem is tackled by implementing data fusion of measurements derived from inertial sensors and custom-developed computer vision algorithms which employ Homography and Optical Flow calculation. With a proper sensorial setup, high-rate and robust ego-motion estimation is achieved, enabling the controlled aggressive maneuverability without reliance on external equipment, such as motion capture systems or Global Positioning System coverage. Finally, a hardware/software framework is developed which adds advanced autonomous perception and navigation capabilities to small-scale unmanned vehicles, employing stereo vision and integrating state-of-the art solutions for incremental environment building, dense reconstruction and mapping, and point-to-point collision-free navigation. Within this framework, algorithms which enable the detection, segmentation, (re-)localization, and mobile tracking –and avoidance– of a dynamic subject within the aerial vehicle’s operating space are developed, substantially increasing the operational potential of autonomous aircraft within dynamic environments and/or dynamically evolving missions.
Open Access
Design, modeling and implementation of open programmable networks
Χαλεπλίδης, Ευάγγελος; Κουφοπαύλου, Οδυσσέας; Κουφοπαύλου, Οδυσσέας; Δενάζης, Σπυρίδων; Γκούτης, Κωνσταντίνος; Κουμπιάς, Σταύρος; Λυμπερόπουλος, Δημήτριος; Βαρβαρίγος, Εμμανουήλ; Κορμέντζας, Γεώργιος; Haleplidis, Evangelos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
Key objectives of network technologies are their overall management and creation of new services. But networks are generally heterogeneous and this creates a problem in their flexibility and integration of new services. Current solutions address the problem piecemeal, without a common framework. The aim of the PhD dissertation is to propose a common framework for resource control and management of a network model alongside with auxiliary and automatic functions to support implementation of new services in open programmable network. This dissertation laid the requirements of the model that would be required and chosen IETF’s ForCES from other framework standards. The dissertation contributed to the ForCES model in multiple ways. It identified gaps and enhanced the current model, made the comparison with similar frameworks and conceptually extended the standard in the resource management plane. Libraries also were proposed and developed, using the model, related to high availability, parallelization and subsidiary mechanisms for installation and activation of new entities. Using this model, the building blocks of a common framework were laid and the architectural structure of open networks was described. Furthermore, this dissertation proposed the conceptual and architectural enhancement of the SDN architecture to include the management plane and the operational plane within devices. With regards to the architecture of NFV, this dissertation proposed to extend the architecture by integrating the interface between the EMS and the NFV using the proposed framework. With regards to auxiliary and automatic features a Domain Specific Language was implemented within this dissertation. The DSL allowed the user to describe the model and then automatically converted that into pieces of code stubs that had implementation of the ForCES standard. The user would only have to fill the required functionality. Finally an original implementation was developed as proof of concept on the LTE/EPC. This dissertation addressed the problems of disproportionate signaling growth and long-distance portions of the EPC architecture using the proposed framework and implemented the management part of the GTP tunneling protocol. In conclusion, the use of a framework for overall management of the entire life cycle of network resources is possible using the proposed advanced ForCES framework.
Open Access
Development of evaluation method for visual design with multivariate statistical techniques
(2013-10-14) Παπαχρήστος, Ελευθέριος; Αβούρης, Νικόλαος; Δερματάς, Ευάγγελος; Ρετάλης, Συμεών; Σγάρμπας, Κυριάκος; Τσέλιος, Νικόλαος; Μαρμαράς, Νικόλαος; Μειμάρης, Μιχάλης; Papachristos, Eleftherios
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The main goal of this thesis is to propose an evaluation method for visual interface design and more specifically for website design. The proposed visual design evaluation method is an adaptation of Preference Mapping (PM) techniques. It is based on overall preference ratings after multiple comparisons of alternative designs and on various multivariate statistical techniques for the analysis, visualization and interpretation of the resulting data. The suitability of the approach for visual interface design evaluation has been explored in four case studies involving overall 149 participants judging 51 websites. In each case study a different website domain was explored in order to examine whether the importance of certain design characteristics is context specific. Heterogeneity in preferences and perceptions was also studied showing that average construct scores are only representative for subsections of the participant sample. In order to aid the preference interpretation process additional data about study websites have been collected from three distinct sources: a) Subjective construct ratings provided by the participants after preference evaluation b) Descriptive attribute ratings obtained from trained expert panel on the same websites c) Objective measures of visual characteristics of the websites In each case study the potential of these types of measurements to explain preference variance has been investigated individually and in combination. The results showed that depending on the characteristics of each case study varying combinations of these types of data had the best explanatory power. A variety of methods (e.g. Internal and External PM) and statistical techniques (e.g. Principal Component Analysis (PCA), Generalized Procrustes Analysis (GPA) and Partial Least Squares (PLS)) have been used in order to summarize and visualize participant preference data of all case studies. In general, the method proposed in this thesis has several advantages over other visual design evaluation methods as for example use of standardized questionnaires. The method is flexible and can be used in various stages of design development but most importantly it allows for the identification of important visual design characteristics without ignoring the diversity that exist both among users and among website domains. These advantages have been demonstrated in the visual design evaluation studies presented in this thesis involving websites from four distinct domains.
Open Access
Development of methodologies for memory management and design space exploration of SW/HW computer architectures for designing embedded systems
(2014-05-16) Κρητικάκου, Αγγελική; Γκούτης, Κωνσταντίνος; Catthoor, Francky; Νικολαΐδης, Σπυρίδων; Κουφοπαύλου, Οδυσσέας; Νικολός, Δημήτριος; Σούντρης, Δημήτριος; Θεοδωρίδης, Γεώργιος; Kritikakou, Angeliki
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This PhD dissertation proposes innovative methodologies to support the designing and the mapping process of embedded systems. Due to the increasing requirements, embedded systems have become quite complex, as they consist of several partially dependent heterogeneous components. Systematic Design Space Exploration (DSE) methodologies are required to support the near-optimal design of embedded systems within the available short time-to-market. In this target domain, the existing DSE approaches either require too much exploration time to find near-optimal designs due to the high number of parameters and the correlations between the parameters of the target domain, or they end up with a less efficient trade-off result in order to find a design within acceptable time. In this dissertation we present an alternative DSE methodology, which is based on systematic creation of scalable and near-optimal DSE frameworks. The frameworks describe all the available options of the exploration space in a finite set of classes. A set of principles is presented which is used in the reusable DSE methodology to create a scalable and near-optimal framework and to efficiently use it to derive scalable and near-optimal design solutions within a Pareto trade-off space. The DSE reusable methodology is applied to several stages of the embedded system design flow to derive scalable and near-optimal methodologies. The first part of the dissertation is dedicated to the development of mapping methodologies for storing large embedded system data arrays in the lower layers of the on-chip background data memory hierarchy, and the second part to the DSE methodologies for the processing part of SW/HW architectures in embedded systems including the foreground memory systems. Existing mapping approaches for the background memory part are either enumerative, symbolic/polyhedral and worst case (heuristics) approximations. The enumerative approaches require too much exploration time, the worst case approximation lead to overestimation of the storage requirements, whereas the symbolic/polytope approaches are scalable and near-optimal for solid and regular iteration spaces. By applying the new reusable DSE methodology, we have developed an intra-signal in-place optimization methodology which is scalable and near-optimal for highly irregular access schemes. Scalable and near-optimal solutions for the different cases of the proposed methodology have been developed for the cases of non-overlapping and overlapping store and load access schemes. To support the proposed methodology, a new representation of the array access schemes, which is appropriate to express the irregular shapes in a scalable and near-optimal way, is presented. A general pattern formulation has been proposed which describes the access scheme in a compact and repetitive way. Pattern operations were developed to combine the patterns in a scalable and near-optimal way under all the potential pattern combination cases, which may exist in the application under study. In the processing oriented part of the dissertation, a DSE methodology is developed for mapping instance of a predefined target application domain onto a partially fixed architecture platform template, which consists of one processor core and several custom hardware accelerators. The DSE methodology consists of uni-directional steps, which are implemented through parametric templates and are applied without costly design iterations. The proposed DSE methodology explores the space by instantiating the steps and propagating design constraints which prune design options following the steps ordering. The result is a final Pareto trade-off curve with the most relevant near-optimal designs. As the scheduling and the assignment are the major tasks of both the foreground and the datapath, near-optimal and scalable techniques are required to support the parametric templates of the proposed DSE methodology. A framework which describes the scheduling and assignment of the scalars into the registers and the scheduling and assignment of the operation into the function units of the data path is developed. Based on the framework, a systematic methodology to arrive at parametric templates for scheduling and assignment techniques which satisfy the target domain constraints is developed. In this way, a scalable parametric template for scheduling and assignment tasks is created, which guarantees near-optimality for the domain under study. The developed template can be used in the Foreground Memory Management step and Data-path mapping step of the overall design flow. For the DSE of the domain under study, near-optimal results are hence achieved through a truly scalable technique.
Open Access
Development of semi-supervised machine learning algorithms and applications
Φαζάκης, Νικόλαος; Fazakis, Nikos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The well-established approach of Supervised learning is a branch of the broader science of artificial intelligence. The aim of this learning philosophy is the development of computer programs to automatically improve their experience through the extraction of useful information from annotated examples. The methodology of this learning approach is extremely useful in real world applications where large collections of data are available related to problems where absolute associations of the input data and the outcomes cannot be discovered or approximated by explicit mathematic formulations. Such scientific fields include observed data of text, audio or image formats. The classic methodology of supervised learning comes with the cost of annotating, usually referred as ‘labeling’ process, the available data instances of a dataset often by human experts in a field. Considering that modern big datasets can have terabytes of data; it is a very inefficient procedure for humans to tackle. This intrinsic bottleneck is addressed by Semi-supervised learning (SSL), which allows the model to incorporate part or all of the available unlabeled data into its supervised learning. The goal of SSL is to maximize a model's learning performance while reducing the amount of labor required by using such newly labeled instances. This thesis is oriented in the improvement of a sub-category of SSL algorithms referred as self-labeled techniques, and the application of them in real world problems. Numerous important questions are answered such as: Which learning algorithms can best utilize the self-labeling schemes? Can the introduction of ensemble learning along with semi-supervised learning provide classification improvements in real world problems such as speaker identification or educational grade prediction? Is it possible to define a new multi-regressor learning scheme based on self-labeling that can rival the existing semi-supervised regression algorithms? Can iterative data imputation be improved through the introduction of self-training? In health-related datasets is it possible to take advantage of unlabeled test sets to balance the shortage of examples through semi-supervised transductive learning?
Open Access
Distributed bandwidth and energy resources’ management in dynamic peer-to-peer video-on-demand systems
Δελτούζος, Κωνσταντίνος; Δενάζης, Σπυρίδων; Κουφοπαύλου, Οδυσσέας; Λυμπερόπουλος, Δημήτριος; Κουμπιάς, Σταύρος; Γκούτης, Κωνσταντίνος; Βαρβαρίγος, Εμμανουήλ; Κορμέντζας, Γεώργιος; Deltouzos, Konstantinos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The rapid development of communications during past decades has made Internet connection available to a vast amount of users which use it for a wide variety of services. Video is one of the most popular services, currently generating the majority of the global Internet traffic and shows growing tendencies for the foreseeable future. As a result, major video-on-demand service providers (i.e. YouTube) have seen their bandwidth costs increase and will inevitably continue to do so. This demand for excessive bandwidth is caused by the simultaneous video watching of large numbers of users. Adopting some peer-to-peer architecture can be a solution to this problem, as users can contribute their resources towards offloading the providers. Indeed, a number of research efforts the past decade have focused on providing live streaming and video-on-demand (VoD) services using peer-to-peer (P2P) architectures. However in these proposed architectures, the dynamic behavior of users has not been sufficiently studied. In a real life scenario, where users can arbitrarily decide to perform a VCR function (stop, fast forward, seeking), ignoring this behavior can significantly deteriorate the system’s efficiency and the perceived QualityofExperience (QoE). Additionally the proposed architectures are designed without taking into account the users’ heterogeneous resources (bandwidth and battery), thus leading to heavily unbalanced systems that fail to utilize these resources. This can severely harm system’s efficiency especially in cases where a significant number of users are mobile nodes with scarce battery resources. In this dissertation, the research area of P2P VoD systems was studied and a distributed system that handles peers’ dynamic behavior and utilizes their resources was developed. Specifically our system is a set of algorithms that optimize the P2P overlay network dynamically and in a distributed fashion, making it adaptive to users’ dynamic behavior and resource changes. The available resources of the participating users are optimally exploited, keeping both the contribution from the media server(s) and the energy consumption to minimal levels. To illustrate the performance of the proposed algorithms, we have developed an extensive P2P VoD simulator that shows the efficiency, scalability and stability of our system under variant and dynamic conditions.
Open Access
Distributed control and optimization of peer to peer video streaming systems
Ευθυμιοπούλου, Μαρία; Κουφοπαύλου, Οδυσσέας; Κουφοπαύλου, Οδυσσέας; Δενάζης, Σπυρίδων; Πάλλης, Ευάγγελος; Κωτσόπουλος, Σταύρος; Λογοθέτης, Μιχαήλ; Βαρβαρίγος, Εμμανουήλ; Κορμέντζας, Γεώργιος; Efthymiopoulou, Maria
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
The purpose of this dissertation is the scalable bandwidth monitoring and control of peer to peer live video streaming systems in order to guarantee the complete and on time video distribution. For this reason, designed, developed and implemented a comprehensive solution that handles the problem of deficit or surplus of bandwidth resources in a peer to peer live streaming system.The major requirements for a P2P live video streaming application can be summarized to the followings: a) efficiency of the video distribution in terms of utilization of peers’ upload bandwidth, in order to minimize any additional bandwidth contributed by a set of media servers (cloud) and/or maximize the playback rate of the video which the system is able to deliver, b) stability of the system which is defined as the uninterrupted and complete stream delivery in each participating peer in the presence of dynamic conditions, c) scalability which is determined by the amount of resources (bandwidth, storage, processing overhead) that the media server/cloud, which manages the system, has to contribute in order to sustain the uninterrupted delivery of the stream, as the number of participating peers grows. There are two strategies in order to adapt the P2P live streaming service to the dynamic upload bandwidth conditions of participating peers. The first is to dynamically adapt the playback rate according to the existing upload bandwidth of participating peers, while the second is to dynamically allocate upload bandwidth from auxiliary sources (e.g. clouds). The selection of a strategy has to do with the QoE that participating users desire and the business model of the service provider. The proposed system is able to guarantee the complete and on time video distribution to every participating peer based on the two aforementioned strategies. The contribution of this dissertation is summarized to the development of these strategies with respect to the aforementioned P2P live streaming requirements. In the first control strategy that has been developed, it is proposed an architecture which adapts dynamically the video playback rate according to the existing upload bandwidth of participating peers. In the second control strategy that is developed, it is proposed an architecture that dynamically calculates and allocates/releases from/to auxiliary sources (e.g. clouds) the amount of deficit/surplus of bandwidth which is required for the video distribution with respect to the video playback rate and system's resources. In the third strategy is proposed a live streaming architecture that with a not only scalable but also in a totally distributed way, is able to determine dynamically the required bandwidth (hence the equivalent in surplus/deficit) for the video distribution. At the same time if necessary subsidiary help, it will no longer come from a central source (e.g. cloud) but from other (additional and/ or unused) peers.
Open Access
Facial feature extraction and estimation of gaze direction in human-computer interaction
Σκόδρας, Ευάγγελος; Φακωτάκης, Νικόλαος; Δερματάς, Ευάγγελος; Μουρτζόπουλος, Ιωάννης; Παλιουράς, Βασίλειος; Μουστάκας, Κωνσταντίνος; Μπερμπερίδης, Κωνσταντίνος; Ψαράκης, Εμμανουήλ; Skodras, Evangelos
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
In the modern age of information, there is a growing interest in improving interaction between humans and computers in an unremitting attempt to render it as seamless as the interaction between humans. In the core of this endeavor are the study of the human face and the focus of attention, determined by the eye gaze. The main objective of the current thesis is to develop accurate and reliable methods for extracting facial information, localizing the positions of the eye centers and performing tracking of the eye gaze. Usually such systems are grounded upon various assumptions regarding the topology of the features and the camera parameters or require dedicated hardware. In the regard of ubiquitous computing, all the methods developed in the scope of the current thesis use images and videos acquired using standard cameras under natural illumination, without the requirement of any special setup and without imposing constraints to the user. The methods developed for the facial feature extraction involve the detection of the lips, the eyebrows and the eyelids. The accurate localization of the eye centers is also extensively studied, comprising the cornerstone on which the study of eye gaze is based. Based on the methods of the previous chapters a system for estimating gaze direction is developed, studying also the different parameters that affect its performance and proposing the optimal setup. Finally, the gaze tracking system proposed is implemented in various practical applications, demonstrating its capabilities and appropriateness for real life applications. All the individual methods proposed have been extensively evaluated in challenging databases and compared to their counterparts found in the literature. The results presented in this thesis indicate the performance superiority of the proposed methods and encourage further research on the field.
Open Access
Fault detection and diagnosis : application in microelectromechanical systems
(2010-12-07T11:59:40Z) Ρέππα, Βασιλική; Τζες, Αντώνιος; Γρουμπός, Πέτρος; Δερματάς, Ευάγγελος; Καζάκος, Δημοσθένης; Κούσουλας, Νικόλαος; Μπερμπερίδης, Κωνσταντίνος; Ποιμενίδης, Τριαντάφυλλος; Τζες, Αντώνιος; Reppa, Vasiliki
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This thesis presents the development of a fault detection and diagnosis (FDD) procedure capable of capturing, isolating and identifying multiple abrupt parametric faults. The proposed method relies on parameter estimation deployed in a set membership framework. This approach presupposes the utilization of a linearly parametrizable model and the a priori knowledge of bounded noise errors and parameter perturbations. Under these assumptions, a data-hyperspace is generated at every time instant. The goal of set membership identification (SMI) is the determination of the parametric set, formed as an orthotope or ellipsoid, within which the nominal parameter vector resides and intersects with the data-hyperspace. The fault detection mechanism is activated when the normal operation of the SMI procedure is interrupted due to an empty intersection of the data-hyperspace and the estimated parametric set. At the detection instant, a resetting procedure is performed in order to compute the parameter set and the data-hyperspace that contain the varied nominal parameter vector, allowing the SMI algorithm to continue its operation. During the fault isolation, consistency tests are executed, relying on the projections of the worst case parametric sets and the ones arisen from the normal operation of SMI. A faulty component is indicated when these projections do not intersect, while the distance of their centers is used for fault identification. In case of the ellipsoidal SMI-based FDD and under the assumption of a time invariant parameter vector, a new fault detection criterion is defined based on the intersection of support orthotopes of ellipsoids. A more accurate estimation of the time instant of fault occurrence is proposed based on the application of a backward-in-time procedure starting from the fault detection instant, while the conditions under which a fault will never be detected by the orthotopic and ellipsoidal SMI based FDD are provided. This dissertation explores the efficiency of the proposed FDD methodology for capturing failure modes of two microelectromechanical systems; an electrostatic parallel-plate microactuator and a torsionally resonant atomic force microscope. From an engineering point of view, failure modes appeared in the microcomponents of the microactuator and the TR-AFM are encountered as parameter variations and are captured, isolated and identified by the proposed FDD methodology.
Open Access
Generalized modeling and stability analysis for modern power systems with large scale integration of converter interfaced RES
Παπαγεωργίου, Παναγιώτης; Papageorgiou, Panagiotis
Τμήμα Ηλεκτρολ. Μηχαν. και Τεχνολ. Υπολογ. (ΔΔ)
This dissertation is placed in the general context of power systems modeling, control, and analysis. Specifically, modern power systems of varying topology featuring large scale integration of renewable energy sources (RES) are considered for the modeling of which, a generalized method is proposed. By taking into account all units located within a modern electricity grid infrastructure, including locally controlled power converter interfaces and electromechanical components, the introduced modeling approach effectively captures the entirety of the system dynamics. Since the proposed formulation retains its structure under changes in grid layout, it can be deployed so as to model varying topology systems, whereas the form of the model itself is nonlinear in order to match and accurately describe the inherently nonlinear nature of modern grid topologies. One of the remarkable properties this approach features is the capability of directly establishing strong stability and state convergence properties for any kind of system that is compatible with this formulation, regardless of its scale or structure. This fact is rigorously proven by adopting a theoretical framework based on advanced nonlinear analysis tools. In this sense, the introduced modern power system representation has a twofold use; it stands both as an accurate and universal power system model and as a direct stability analysis tool. An important part of this dissertation is also devoted on developing suitable control schemes for various types of power converter interfaces, mainly focusing on the one type dominating the power grids nowadays, the three-phase voltage source converter (VSC). The proposed control designs are of simple PI-type and they feature several innovations in their structure that effectively withdraw many adversities this kind of control schemes usually have. In this frame, several VSC-based configurations are considered, including the cases of a VSC connected to a stiff or a weak grid. For the latter case, a novel PLL mechanism is also proposed that effectively copes with the challenging issue of maintaining synchronism. In addition, VSC-based dc-link configurations are considered and appropriate control designs are proposed for such topologies as well. The stabilizing effect of the developed control designs on the closed-loop performance of these systems can be proven by adopting the theoretical framework used in the unified CIAT model analysis. As it becomes apparent, all of the considered topologies are absolutely compatible with a simpler variation of the CIAT model and thus, strong stability and state convergence properties can be directly extracted. The case of constant power loads (CPLs) is also investigated in relation to the developed CIAT representation. Several aspects regarding the need of characterizing some loads as constant power ones in the modern power system paradigm is also discussed, whereas the basic properties of such loads are examined by considering a typical dc/dc boost converter/CPL configuration. Finally, the complete CIAT formulation is employed in order to accurately model a realistic modern DG-based power system example featuring both controlled power converter interfaces and electromechanical components. In this process, the remarkable properties accompanying this modeling approach are being fully showcased.