22C:161 Introduction to Robotics Spring 1995

Homework 4 - Due Monday , February 27

In this homework assignment, you will implement robot inverse kinematics computations corresponding to the material covered in Chapter 4 of the book. Specifically, you will write routines that will read compute inverse kinematic solutions for specific robots. The routines will be based on analytic inverse kinematic solutions. Though possible for some classes, we will not implement numerical techniques that would allow us to take any robot specification as input (like in Homework 3) and compute inverse kinematic solutions (recall that although it is not even possible, in general, to write a routine that can take an arbitrary robot as input and and compute all its inverse kinematic solutions, it is possible to write a routine that will find, for arbitrary robots, at least one of its inverse kinematic solutions).

1. Implement routines that compute the inverse kinematics solutions for the robot of example 3.3 (this is one you did the forward kinematics for on the previous homework, so you should have a robot specification file for it). Thus, from three goal inputs, x, y, and phi, your routines should calculate all the sets of joints angles, theta1, theta2, and theta3 that achieve the goal.
2. Implement routines that compute the inverse kinematics solutions for the Puma.
3. For any input goal, display one of the solutions graphically. You can do this in a variety of ways - (1) print a list of the possible solutions and ask the user to choose one, (2) implement a routine that selects the "closest", in some sense that you define, (3) cycles through different solutions as you click a button or something, etc.
4. Write routines corresponding to MOVE and MOVES on our robot. For MOVE, you'll need just one call to the inverse kinematics routines. For MOVES, you'll need to call them many times (you'll get an approximately straight line).