LEARNING KINEMATICS FOR ROBOTS WITHOUT USING TRANSFORMATION MATRICES: A COMPUTER-BASED APPROACH

摘要

Study of the kinematics of robots, especially the inverse kinematics, generally requires knowledge of a very high level of linear algebra. This is usually necessary because transformation matrices are used to describe robot position, relate positions of individual links to the position of the end-effector and describe inputs such as a set of all input angles. Due to this mathematical requirement, the opportunity to study this technical subject has been traditionally limited to a small group of senior undergraduate or graduate level students pursuing courses in robotics. In this paper, an attempt to implement this technical subject in an introductory Kinematics of Machines course for students pursuing B.S. Manufacturing and Mechanical Engineering Technology degrees is described. Students in this course need only to have knowledge of elementally algebra, trigonometry and calculus but not higher level linear algebra. In the kinematics of robots module, students are introduced to a simple 2-link SCARA-type robot called the Telesis robot. Due to its planar configuration, the forward and inverse kinematics of the Telesis robot can be analyzed using only simple algebra and trigonometry. With this simple level of mathematics however, it is still possible to introduce important robotics concepts such as workspace, singularity, and branching.