AbstractThis paper discusses the trajectory planning problem of a four-degrees-of-freedom (4-DOF) robotic manipulator with similar upper limb structures of human beings. The spatial motions of the end-effector (i.e., wrist) can be desired in terms of controlling four motors of the robotic manipulator; hence, such a configuration results in the redundancy problem. In general, Jacobian solutions perform linear approximations of inverse kinematics problems with redundancy conditions. With human beings, the same wrist position can be found from different limb postures, and these postures depend on different motion scenarios such as writing words, waving hands and shaking hands. Nevertheless, Jacobian solutions are hardly to realize specific limb motion scenarios of human beings. Therefore, this paper proposes a supervised neural network based robotic manipulator trajectory planner which constructs limb motion characteristic models according to relative joint posture features with respect to different motion scenarios. The motion features are further used to provide an auxiliary condition for eliminating the redundancy problem of the inverse kinematics as well as to meet specific motion scenarios. Finally, several trajectory planning results were evaluated in terms of the Jacobian and neural network approaches respectively by using an optical motion capture system. Experiment results demonstrated that the proposed neural network based control system performs similar motion behaviors when compared to Jacobian approaches based on the same test trajectory and motion scenario.