Abstract: To solve the problem of gait trajectory planning during the rigid operation of lower limbs, we propose an online gait trajectory planning method and control strategy based on dynamic movement primitives. The technique uses plantar pressure to segment the gait phase, a linear inverted pendulum model to determine the location of the landing site, inverse kinematic programming to determine the angle of each joint, and a dynamic movement primitives algorithm to learn online the trajectory of each previous gait phase. We apply the previously obtained joint angle as the target endpoint for real-time gait trajectory planning. Finally, we control the exoskeleton according to the control strategy from the generated curves. The algorithm is verified using a 6-DOF exoskeleton, and the results show that the proposed algorithm can effectively learn and generate the gait trajectory online according to the transformation of the gait phase, ensuring a certain degree of motion stability.