Recently, a prototype of a hand exoskeleton for post-stroke rehabilitation purpose was proposed by the Group of Robotics, Automation and Articular Biomechanics (GRAB) at the Department of Industrial Engineering, University of Bologna. The prototype comprises five planar mechanisms (one per finger) globally actuated by two DC motors. A total of fifteen human-machine connections are needed to fasten the device to the patient hand. The moving link of the thumb mechanism is actuated by a spatial RSSR mechanism whose frame link geometry must be ad hoc regulated every time the device is fitted on the patient hand. With the future goal to build a new version of the hand exoskeleton, in this dissertation three problems arising from this prototype were tackled. The first problem regards the need to lower the number of human-machine connections needed to fasten the exoskeleton to the patient hand. A new finger mechanism that permits to lower the total number of human-machine connections from fifteen to only six was proposed. The second problem regards the synthesis of the RSSR mechanism. A novel synthesis procedure was proposed in order to guarantee the optimal motion and force transmission to the thumb mechanism once the hand exoskeleton is fitted to a new patient, i.e. for different frame link geometries of the RSSR mechanism. The third problem regards the need to approximate the finger phalange motion as a rotation about a revolute axis. In this perspective, two different joint axes identification techniques were proposed. The techniques are based on the Burmester theory (a theory generally used for the synthesis of mechanisms), here used in an original way to identify an axis of rotation. A comparison of this two technique with a more standard technique based on the finite helical axis is reported.

The knee joint plays a central role in human motion for its dual function: providing a large range of motion in flexion/extension and stability in the other degrees of freedom. Computational modeling is a powerful tool to deepen our understanding of the joint mechanics, overcoming the main limitations of experimental investigations, i.e. time, cost and impracticability, and providing valuable insights for prosthetic design, rehabilitation and surgical planning. Within this background, the specific aim of this dissertation is threefold: to develop a sequentially-defined kinetostatic model of the knee, comparing the performance of spherical and anatomical surfaces; to develop a dynamic model of the knee to predict the quadriceps force during the squat activity; to estimate the compressive force that the implanted knee joint needs in order to reproduce natural stability. This dissertation presents novel and efficient procedures to model and evaluate the behavior of the natural and implanted knee under the effect of static and dynamic loading conditions, extending the current knowledge in the field of musculoskeletal computational modeling.

Silicon-based power devices have dominated power electronics applications over the last decades. Research and development in microelectronics have pushed the performance of power devices to face some fundamental limitations of silicon material. Wide band-gap semiconductors, such as silicon carbide, offer a solution to the pressing energy efficiency performance requirements of power electronic systems. Silicon carbide power devices can operate at higher temperatures, higher frequencies, and generate less power losses as compared to traditional silicon-based technologies. The use of wide band-gap transistors, however, is not the only way to increase the efficiency of the converters. Special DC to AC topologies, named soft switching converters, can be adopted as well in order to reduce the switching losses of transistors. The development of DC to AC power converters for applications requiring high efficiency is presented in this thesis. Silicon and silicon carbide based inverters, as well as soft switching inverters, have been analyzed and fabricated for performance comparison.

The knee is one of the most complex and studied joint of the musculoskeletal system provided its great importance in locomotion. Therefore, a deep understanding of its behaviour and of the role played by each of the structures composing it is fundamental. Knee joint models are an invaluable tool to understand the behaviour of the knee and their usefulness is proved in many fields such as surgical planning and prosthetic design. A huge amount of models has been proposed in the literature focusing on the kinematic, the kinetostatic and the dynamic behavior of the joint. Models can be based on in vivo or in vitro data. While the kinematic and the kinetostatic models are defined properly on in vitro data, the dynamic ones cannot. This discrepancy leads to a gap, a lack of coherence, between the usually in vitro defined kinematic and kinetostatic models and the study of the active structures of the joint. In order to achieve a comprehensive knee joint description in which the kinematic, kinetostatic and dynamic models coherently stem one from the other, the identification of a procedure that allows to obtaining reliable kinematic and kinetostatic models in vivo is needed. In the present dissertation a procedure is defined that allows for the identification of a subject specific knee joint model in vivo starting from standard clinical data obtained by the use of non invasive techniques such as computed tomography (CT), magnetic resonance imaging (MRI) and fluoroscopy. This procedure leads to an accurate identification of the parameters needed to personalize the 5-5 parallel mechanism and its patello-femoral extension on a single patient in order to accurately reply the knee joint original motion. Furthermore, following the sequential approach to the modelling of the joint, a stiffness model of the knee is specialized on the specific subject's anatomy.

The present dissertation aims to explore, theoretically and experimentally, the problems and the potential advantages of different types of power converters for “Smart Grid” applications, with particular emphasis on multi-level architectures, which are attracting a rising interest even for industrial requests. The models of the main multilevel architectures (Diode-Clamped and Cascaded) are shown. The best suited modulation strategies to function as a network interface are identified. In particular, the close correlation between PWM (Pulse Width Modulation) approach and SVM (Space Vector Modulation) approach is highlighted. An innovative multilevel topology called MMC (Modular Multilevel Converter) is investigated, and the single-phase, three-phase and "back to back" configurations are analyzed. Specific control techniques that can manage, in an appropriate way, the charge level of the numerous capacitors and handle the power flow in a flexible way are defined and experimentally validated. Another converter that is attracting interest in “Power Conditioning Systems” field is the “Matrix Converter”. Even in this architecture, the output voltage is multilevel. It offers an high quality input current, a bidirectional power flow and has the possibility to control the input power factor (i.e. possibility to participate to active and reactive power regulations). The implemented control system, that allows fast data acquisition for diagnostic purposes, is described and experimentally verified.

No description available.

No description available.

No description available.

No description available.

Machines with moving parts give rise to vibrations and consequently noise. The setting up and the status of each machine yield to a peculiar vibration signature. Therefore, a change in the vibration signature, due to a change in the machine state, can be used to detect incipient defects before they become critical. This is the goal of condition monitoring, in which the informations obtained from a machine signature are used in order to detect faults at an early stage. There are a large number of signal processing techniques that can be used in order to extract interesting information from a measured vibration signal. This study seeks to detect rotating machine defects using a range of techniques including synchronous time averaging, Hilbert transform-based demodulation, continuous wavelet transform, Wigner-Ville distribution and spectral correlation density function. The detection and the diagnostic capability of these techniques are discussed and compared on the basis of experimental results concerning gear tooth faults, i.e. fatigue crack at the tooth root and tooth spalls of different sizes, as well as assembly faults in diesel engine. Moreover, the sensitivity to fault severity is assessed by the application of these signal processing techniques to gear tooth faults of different sizes.