A parameter‐identification‐based diagnosis for detecting static air‐gap eccentricity

A new method for predicting static eccentricity in an induction machine with a non‐uniform air gap is described. This procedure is based on a suitable system mode I for which the model parameters have to be identified. Under certain conditions, the mathematical model of the machine leads to an equation depending on speed, voltage, current and their derivatives. The parameters of the machine can be identified by a least‐square algorithm. The basic concept of the identification method is that the mathematical parameters will be influenced by the static eccentricity because a non‐uniform air gap will increase the inductance in all three phases of the machine. Practical results of air‐gap eccentricity detection of a low‐power motor are presented.