Modelagem, análise e detecção de assimetrias de rotor em máquinas de indução.

• Ano de defesa: 2006
• Autor(a) principal: Carla Cesar Martins Cunha
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Engenharia Elétrica; Engenharia elétrica
• Link de acesso: http://hdl.handle.net/1843/BUOS-8CSGYG

Rotor bar and end-ring faults yield asymmetrical operation of squirrel cage induction machines causing unbalanced currents, torque pulsation, increased losses and decreased average torque. The need of proper characterization of motor operation performance under these conditions, and the detection of rotor faults at an earlier stage, has pushed the development of monitoring methods with increasing sensitivity and noise immunity. Equally important is the analysis and quantification of how these asymmetries affect the motor performance. Essential contributions to such effort are: the modelling of the induction machine including rotor bar and end-ring faults to any extent, with minimum computational complexity, and that machine parameters required for these models are easily determined. Therefore, a new modelling technique based on the dq model of the induction machine is presented in this work. The model is always of the fifth order, independent of the number of rotor bars, and the rotor current transformation matrix only depends on the number of rotor bars and can be computer generated. Several rotor faults situations were investigated employing computational simulations. In order to verify experimentally the proposed modelling technique, a special induction motor, named Frank, was constructed to be used in this work. A 10cv standard induction motor was also employed, besides Finite Element Analysis. Stator motor current frequency spectrum, rotor bars RMS currents and the Vienna Monitoring Method are included to support the discussions and the proposed approach. A novel technique for inverter fed squirrel cage induction motor broken bars detection, based on the current regulators error signals already present in the drive, is also proposed in this work.