Robust swing leg controller under large disturbances

Abstract

Local swing leg control was developed that takes advantage of segment interactions to achieve robust leg placement under large disturbances while generating trajectories and joint torque patterns similar to those observed in human walking and running. The results suggest the identified control as a powerful alternative to existing swing leg controls in humanoid and rehabilitation robotics. Alternatively, a detailed neuromuscular model of the human swing leg was developed to embody the control with local muscle reflexes. The resulting reflex control robustly places the swing leg into a wide range of landing points observed in human walking and running, and it generates similar patterns of joint torques and muscle activations. The results suggest an alternative to existing swing leg controls in humanoid and rehabilitation robotics which does not require central processing.

1 . A model-based limb controller for a limb comprising at least one robotic limb joint, the controller comprising:
 a non-neuromuscular model including a swing leg model and dynamic equations, wherein the non-neuromuscular model is configured to receive feedback data relating to a measured state of the limb and, using the feedback data and the swing leg model and the dynamic equations to determine at least one torque command to be applied to the at least one robotic limb joint; and   a torque control system in communication with the non-neuromuscular model, wherein the torque control system receives the at least one torque command from the non-neuromuscular model for controlling the at least one robotic limb,   whereby motion of the at least one robotic limb joint is determined without enforcing reference trajectories of the at least one robotic limb joint.   
     
     
         2 . The controller of  claim 1 , further comprising at least one sensor mounted in proximity of the at least one robotic limb joint to provide the feedback data to the at least one robotic limb joint. 
     
     
         3 . The controller of  claim 1 , wherein the at least one robotic limb joint is only an artificial knee joint. 
     
     
         4 . The controller of  claim 1 , wherein the at least one robotic limb joint is an artificial knee joint and an artificial hip joint. 
     
     
         5 . The controller of  claim 1 , wherein at least one sensor is selected from the group consisting of an angular joint displacement, a velocity sensor, a torque sensor, and an inertial measurement unit. 
     
     
         6 . A method for controlling a limb comprising at least one robotic limb joint, the method comprising the steps of:
 receiving feedback data relating to a state of the limb;   determining at least one joint torque to be applied to the at least one robotic limb joint based on the state of the limb using a non-neuromuscular model comprising a swing leg model and dynamic equations; and   applying the at least one joint torque determined by the non-neuromuscular model processor at the at least one robotic limb joint,   whereby motion of the at least one robotic limb joint is determined without enforcing reference trajectories of the at least one robotic limb joint.   
     
     
         7 . The method according to  claim 6 , wherein the at least one robotic limb joint is only a robotic knee joint. 
     
     
         8 . The method according to  claim 6 , wherein the at least one robotic limb joint is a robotic knee joint and a robotic hip joint. 
     
     
         9 . The method according to  claim 6 , wherein the state of the limb is selected from the group consisting of a current hip rotation angle (φ h ) and a current knee rotation angle (φ k ).