The biped is controlled using Dynamixel A12 and/or A18 motors. We have created a Matlab interface for the Dynamixel motors. This interface enables controlling the motor actuation, as well as reading motor values and parameters. There is even an associated trajectory class that takes in trajectories and can replay them.

The first step here is to get to know the biped, as well as to model the biped that you will be using. Depicted below is a sketch of the biped model. Go meet up with the biped contact and get started with the motor interfacing, plus the biped modeling.

1. You should be able to move the motors as well as read their values. In addition, you should be able to describe how to calibrate the motors, or what is involved in the construction of the biped to ensure proper calibration. Do so in your homework submission.

2. Measure the link lengths for your biped and return the values $l_0$ to $l_4$. If there is no $l_0$ value yet, then use something sensible and state as much. Using this value for now. Later, it will get locked down.

3. Using the picture below, turn in the forward kinematics that describe the left foot frame relative to the torso frame, as well as the right foot frame relative to the torso frame. There should be a function written in your biped class that is for the torso frame forward kinematics. It returns two frames $g^T_{LF}$ and $g^T_{RF}$ for th left/right foot, respectively.

4. Flesh out the plotTF member function of the biped class. It plots the robot relative to the torso frame. As long as the torso frame is not set to anything, then it should plot at the origin (meaning the torso frame is the identify configuration). The predicted, forward kinematics left and right foot frames should align with the feet for whatever joint angles are chosen for the manipulator. In essence, executing this function, together with plots of the reference frames should give a nicer version of the picture below.

Back – Main