Automatic Identification of Axle center Trajectory Shape Based on Invariable moment Characteristics and a New Type of Correlativity

A novel approach for the automatic identification of axle center trajectory shape is presented. First, the vibration displacement signals acquired in two mutually vertical directions were through a noise suppression treatment and a fitting process made to form an axle center trajectory. All enclosed zones were filled with borderline colors, resulting in the formation of an axle center trajectory image. The invariable moment characteristic vectors were calculated. Then, with the invariable moment characteristic vectors of the various axle center trajectories generated by simulation serving as reference modes and those actually detected serving as the modes to be identified, an identification was carried out with the help of correlativity. To enhance the precision of mode identification, proposed was a high resolution correlativity calculation method. By use of the latter the axle center trajectory shape of model MJF 30 6 electric generator was identified under three operating conditions, i.e., no load operation, on line normal operation, rotor inter turn short circuit fault based operation.