Online compensation of detuning effects in vector control based on adaptive sensorless rotor thermal model identification

Abstract

Temperature-related variations of the rotor resistance induce detuning effects in a vector control drive. This paper focuses on the development of an adaptive sensorless rotor thermal model identification algorithm to compensate for such detuning effects. The proposed rotor thermal model identification algorithm is divided into two parts: a sensorless rotor temperature estimator and an adaptive rotor thermal model identifier. First, the rotor temperature estimator utilizes a sensorless rotor speed detector and an online rotor resistance estimator to estimate the rotor temperature. This estimated rotor temperature serves as a reference signal to the subsequent adaptive identification of rotor thermal model parameters. Then, based on the analysis of the rotor thermal model, a digital antialiasing filter using a Kaiser window is designed to suppress the noise in the previously estimated rotor temperature in conjunction with a downsampling stage. After that, the rotor thermal model parameters are identified online to establish the relationship between the motor losses and the rotor temperature. Once these thermal parameters are successfully determined, the rotor temperature can be predicted from the motor losses alone, in an online fashion, and such predicted rotor temperature can be used to provide proper compensation to the detuning effects. The proposed algorithm is validated through experimental results.