A feedback control strategy based on Kalman filter was proposed to quantitatively study the drive train torsional vibration mitigation of doubly-fed wind turbine, and the control effect of drive train torsional vibration was compared through simulation calculation. Taking the drive train of 7.0 MW doubly-fed wind turbine as the research object, the Kalman filter was used to estimate the twist angle of the drive train, and an additional electromagnetic torque of the generator was designed for torque control based on the estimated torsional speed of low-speed shaft. The load and power generation calculations were compared with virtual damping control and no-damping control over 20 years full life cycle. The results show that, the correlation between the twist angle of the low-speed shaft estimated by the Kalman filter and the actual value can reach 0.99. The key differences between the feedback control based on the Kalman filter, the virtual damping control and the non-damping control are as follows. The equivalent fatigue load of the low-speed shaft of the drive train reduces by 2.11% and 4.89%, respectively. The equivalent fatigue load of the high-speed shaft of the drive train reduces by 1.99% and 4.78%, respectively. The power generation reduces by 200 kW·h and 700 kW·h, respectively. It can be concluded that the Kalman filter has a good estimating effect on the twist angle of drive train, and the designed additional electromagnetic torque based on the estimated torsional speed of low-speed shaft obtained by Kalman filtering has a very good suppression effect on the torsional vibration of the drive train.