On-line Identification of Induction Machine Rotor (Parameters) in Speed Sensorless Operation based on Saliency Extraction

This study presents a new online rotor resistance/temperature detection technique for the utilization in speed sensorless machine control. Accurate estimation of induction machines rotor resistance is very important to ensure high dynamic operation as well as precise torque production. The technique proposed is based on analysis and exploitation of inherent machine saliencies. More particularly, the angle of the saturation saliency can be used to monitor the rotor resistance. For exploitation of the machine saliencies, a transient voltage pulse excitation method was accomplished. The operating point dependent properties of the saturation saliency are identified during commissioning of the speed sensorless control. Then properties that are independent of temperature are used as a reference for the rotor resistance identification. A signal processing structure is set up to tune a fundamental wave model of the machine thus identifying the actual parameter value of rotor resistance. This structure is thus able to track the parameter value under low speed operation and changing rotor temperature. Measurement results are presented to verify the accuracy of the proposed method by comparing estimated and sensor-based values of torque, speed, slip and temperature.