Rotor Displacement Detection of Magnetic Bearing by Using Gradient Descent Algorithm with Differential Coil Currents

To improve the integration level and reduce the cost due to the need of individual displacement sensors for active magnetic bearing (AMB) system, the self-sensing of rotor displacement is a promising technique and need further improvement in the detection accuracy. In this paper, an improved rotor displacement detection method is proposed by using the gradient descent algorithm with differential coil currents. Firstly, the instantaneous slope of coil current ripples for a two-phase three-bridge switching power amplifier is accurately measured for self-sensing. Secondly, two-phase currents of differential coil in the same degree of freedom (DOF) are synchronously sampled to double the number of data points, and the algorithm gradient descent algorithm is included to realize the self-sensing displacement detection. Finally, the suspension experiment based on self-sensing control is carried out on the test platform of 4-DOF AMB platform, and the effectiveness of the proposed method is verified.