Trajectory control of a bipedal robot using feed forward compensation methodology

Abstract

In this survey, trajectory control of lower limb joints of a 2D walking robot is introduced by using feed forward compensation control methodology. The mathematical model of walker is considered as a combination of two serial manipulators, each having two revolute joints, in other words, having two degrees of freedom. Inverse kinematics analysis and recursive Newton-Euler computation methods are used to obtain the dynamic equations, which describe the motion of the walking system. For desired walking characteristics, hip and ankle trajectories are derived. Actuators used in the system are permanent magnet direct current motors and their state space representation is given. Finally, feed forward compensation architecture is introduced to control hip and ankle trajectories and simulation results are given. These results show that feed forward compensation method can be used efficiently for the solution of tracking problem of bipedal walker in order to obtain a natural way of walking in both stance and swing phases.