Analytical Solution for the Inverse Kinematics Problem of the Franka Emika Panda Seven-DOF Light-Weight Robot Arm

摘要

The redundant light-weight robot Franka Emika Panda is omnipresent in modern research like human-robot interaction or machine learning. Much of that research requires to solve the inverse kinematics problem (IKP) properly. This paper presents an analytical solution for the IKP based on the geometry of the manipulator's kinematic structure. It also provides information on how the redundancy is resolved and how singularities are handled. The whole procedure is a concatenation of intersection and closest point problems of two or three dimensional objects, for which at least basic solutions already exist. Consequently, the approach is lightweight in calculation and, in addition, very robust when it comes to finding solutions for rare edge cases. On the top of that, the simple nature of calculations make the approach utilizable in a real-time safe fashion. Hence, an implementation in C++ is given as open source library, which also can be used to validate the results of this paper.