ABSTRACT

This chapter presents a linear solution for the simultaneous identification of both the BASE and TOOL transformations. Many robotic applications require accurate knowledge of the geometric inter-relationships among the robot, its sensors and a reference frame. The base coordinate frame is normally defined inside the robot structure. Its default location is often provided by the robot manufacturer; however users are allowed to redefine it. The flange coordinate frame is defined on the mounting surface of the robot end-effector. To simulate an actual manipulator, uniformly distributed random noise processes were also injected to the PUMA joint variables and link parameters. The linear approach for robot simultaneous BASE and TOOL calibration by solving homogeneous transformation equations of the form A X = Y B was first proposed by Hanqi Zhuang, Zvi S. Roth and R. Sudhakar. The error-model based technique has the advantage of being easy to implement, but it needs initial conditions.