LAUSR

Abstract Axial unbalance of the rotor is an important origin of disturbance torque for the GyroWheel (GW), which leads to constant torque and measurement drift. This paper proposes a novel axial unbalance identification strategy based on multi-position calibration. Based on the gimbal assembly support characteristics of the tilting rotor, the disturbance torque model caused by the offset from the support center to the centroid is established. Combined with multi-position calibration, the least square method is introduced to estimate the drift error coefficients, which are taken as the equivalent vibration response. The relationship between the unbalanced excitation by adding trial weight and the vibration response helps to calculate the original unbalance. To avoid the non-stationarity of the current signal affecting the estimation accuracy of the drift error coefficients, a signal process with consideration of CEEMDAN-IIT denoising is exploited. The experiments are performed to verify the superior performance of the proposed identification method.