Abstract: This paper presents a large-scale Stewart platform-based six-component force/torque sensor.A force/torque measurement model is established and an optimal objective function based on the inverse kinematics is offered.A novel parameter calibration method,named branch-in-turn calibration,is presented.Based on the developed prototype of the six-component force/torque sensor,experiments have been made.The experimental results show that the branch-in-turn calibration method can effectively identify the structural parameters of the sensor and the measurement accuracy is improved. This method can be applied to parameter calibration of the six-component force/torque sensor with similar structure.