Abstract:In the humanmachine interaction (HMI), the sensor signal acquired by the lower extremity exoskeleton of the human body lags behind the actual form of the human body motion, thereby causing the exoskeleton robot to fail to follow the motion of the hip joint and the knee joint in real time, and there is a problem that no assistance can be provided. A lower extremity exoskeleton model is designed. Based on which, the force/torque sensor to obtain the signal, the use of Kalman filter algorithm to obtain the information obtained by the motion prediction, and then input the prediction signal to the lower extremity exoskeleton on the simplified twolink kinematics model of the robot model, the stability of the system and the accuracy of the prediction were checked by using ProportionDerivative (PD) control. Finally, the MATLAB simulation experiment results show that the signal can be used to effectively predict the form of human movement after Kalman filtering. The exoskeleton leg swing can effectively follow when the human body is moving, thus making up for the delay.