Test-bed system for improved induction machines diagnostics

Diagnostic techniques for induction machines based on the Motor Current Signature Analysis monitor the increase of specific frequencies of the stator current spectrum, that are directly linked to anomalous conditions. The stator current spectrum includes several lines, that need to be accurately analyzed in order to achieve a reliable fault diagnosis. Several papers in the literature deal with the analysis of stator currents lines, while it is hard to retrieve information on rotor cage variables. In this paper a dedicated induction machine and experimental set-up developed at IEN is used, that allows the acquisition of instantaneous bar currents, through current sensors mounted on the rotor bars. Though artficial faults are not not introduced in the machine, intrinsic non-idealities introduce harmonic distortion in the machine variables. Rotor instantaneous currents are computed by finite element approach and the results are compared with the experimental ones. Then the FFT algorithm is applied to scalar and complex variables and the results are analyzed with the help of the rotating field approach. Reference is made to synchronous operating conditions in order to address the rotor current components caused only by non-idealities. ©2005 IEEE.