Design, Modeling and Diagnostics of Closed-Loop Inverter-Fed Electrical Machines

This thesis aims to present the main research activities carried on by the author during his doctoral studies. Two main topics have been studied during this period: the design and modeling of synchronous brushless machines and the diagnostics of closed-loop controlled induction motors. The design of Interior Permanent Magnet (IPM) machines, nowadays, has a crucial role in developing the whole drive. After reviewing the literature already published, two design procedures will be presented in this thesis. As will be explained in the first part of this thesis, the design of a machine is strongly influenced by the requirements and the application. An important focus will be given to the behavior of the machine in the field-weakening region, implementing models that also take into account the thermal and mechanical limits of the machine under certain conditions. More in detail, an analysis of the reduction of the torque ripple by adopting specific skewing techniques will be provided. Moreover, a detailed study of the costs of different materials for the manufacturing of this kind of machine will be given. \noindent Induction motors represent the most widely used rotating electrical machine typology. In the second part of this thesis, based on a literature review of the topic, a method for the detection of the broken rotor bar fault while the motor is operating under closed-loop conditions based on the Motor Current Signature Analysis (MCSA) will be proposed. The key point behind the detection of fault under these operating conditions is the problem that arises from the non-constant frequency of the fundamental harmonic even when the start-up transient is already depleted. Moreover, the control loop operation introduces a high noise level in the signals, which increases the difficulties in the identification of the harmonics. Several signal analysis techniques will be presented, together with their theoretical background. Signals from the phase current of healthy and faulty induction motors operating in different control conditions will be analyzed using traditional methods such as the fast Fourier transform and the short-time Fourier transform. The same signals are analyzed through the Dragon Transform and the Short-Time Min-Norm. These methods ensure high resolution in frequency and time, allowing the detection of trajectories of the harmonics in all the analyzed conditions.