An Algorithm for the Solution of Inverse Kinematics Problems Based on an Interval Method

In this paper we describe a general procedure to solve the positional inverse kinematics problem using an interval method. The classical interval Newton method is used together with specific developed interval cuts over the closure equations. The algorithm uses a basic branch-and-bound procedure to obtain all solutions of the inverse positional problem of arbitrary single-loop kinematic chains. Some preliminary examples are given, although still much improvement can be done in the heuristics involved and the code optimization, leaving plenty of room for future improvements.