A Hybrid Appproach to Inverse Kinematics Modeling and Control of Pioneer 2 Robotic Arms

— For robotic manipulators that are redundant or with high degrees of freedom, an analytical solution to the inverse kinematics is very difficult or impossible. As alternative approaches, neural networks and optimal search methods have been widely used for inverse kinematics modeling and control in robotics. This paper presents a first analytical solution to the inverse kinematics of a widely used robotic arm (Pioneer 2 robotic arm), which, combined with an optimal search method, provides an effective solution to the modeling and control of the Pioneer 2 robotic arm. I. INTRODUCTION Inverse kinematics modeling has been one of the main problems in robotics research. The most popular method for controlling robotic arms is still based on look-up tables that are usually designed in a manual manner [1]-[3]. Alternative methods include neural networks [4]-[11] and optimal search [12], which often encounter problems caused by the fact that the inverse kinematics systems of most robotic arms are multi-valued and discontinuous functions [9]. For robotic manipulators that are redundant or with high degrees of freedom (dof), there are hardly effective solutions to the inverse kinematics problem except for the manually designed look-up table method that is limited to applications with a priori known trajectory movements. The Pioneer 2 robotic arm (P2Arm) developed by ActivMedia Robotics has been widely used for robotics research, teaching, and development (http://robots.activmedia.com/). However, to date there is no analytical inverse kinematics solution for the P2Arm.