Extension of Inverse Kinematic Solution for a Robot to Cope With Joint Angle Constraints

In this study, inverse kinematics solution based on numerical methods that can be applied universally to various robot systems was introduced. In particular, it was solved by analyzing and solving various problems of the existing inverse kinematics method from different perspectives by constructing and solving this problem as a mathematical optimization problem rather than the general numerical inverse kinematics method obtained using Jacobian. In addition, to verify the efficiency and stability of the proposed method, various simulations and tests were performed by applying to bin picking process using the 5-and 6-axis robot. Finally, in case the geometric information of the target object that was obtained based on deep learning was given, the inverse kinematics solutions for the robot led the robot to pick it up and move it to the target point.