Smooth joint trajectory planning for humanoid robots based on B-splines

To guarantee the walking stability of humanoid robots, a new joint trajectory planning method is proposed. First, to map the trajectory of task space into the trajectory of joint space, a numerical algorithm of inverse kinematics is used to obtain the time sequences of joint angle. In this algorithm, the return values calculated by forward kinematics are compared with the expected pose repeatedly until the error is less than the threshold value. Then seven-degree B-splines are exploited to generate smooth joint trajectories with continuous velocity, acceleration and jerk. By representing kinematics constraints of humanoid robot with control points of B-splines, the motion parameters of both endpoints of the trajectory can be set to desired values. Simulation results show that by using the proposed joint trajectory planning method, smooth trajectories with continuous jerk can be obtained and these trajectories can be used to generate stable walk motion.