Abstract
In this paper, we propose a new nonlinear programming based inverse kinematics solver for generating various motions of a multiple articulated robot. Our method utilizes L-BFGS-B method which can handle inequality constraints intrinsically and whose computational cost is low. Our method can generate various motions of a multiple articulated robot only by setting up objective functions. In this paper, we provide several objective functions one of which can generate the motion of a floating-base robot and which can generate self-motion of the redundant robot without calculating the pseudoinverse of the Jacobian matrix. Through computer simulation, we confirm the validity of our method.
