Видео

Hello Professor, I was trying to track a circular trajectory and avoiding a collision from an obstacle placed at some position on the circular trajectory. So, combining the V(q) and h(q) and then re-arranging the terms to put it in Ax <= b form would work ? I am trying to do it, but for some reason, when the robot is coming close to the obstacle, then it is showing some weird behavior. Does this has to do with the last terms in V(q) and h(q). The last term in V(q) implies: The last term in the control Lyapunov function, is introduced to steer the unicycle towards the goal, i.e., to align its longitudinal direction with the line segment connecting the point to the goal. Further, the last term in h(q) implies: The last term in the control barrier function, is introduced to steer the unicycle away from the obstacle, i.e., to align its longitudinal direction with the line segment connecting the point to the obstacle. PS: I am not using the relaxation term, just min || u || ^2

Hello Professor, I tried the collision avoidance while going to a desired point using the technique you told and also the code you shared via google colab. I have one doubt. For sure, if we change the size of the obstacle of the safety margin, so in your code it was uni.set_avoid_pose([7.5, 5.0], 0.75), where this 0.75 was the value of D (kind of safety margin). So if we change this D, the plot of h will also change. But for any value of D, do you think that the plot might be negative? I tried with an higher value =2, and for some instances h became negative. So is there any limitation of this approach or do we have any constraint on the safety margin (that it cant be greater than a particular value). A comment from you would be appreciated. Thanks

Hi, is there any video for Lecture 1 as well?

how to make a video like this！

Nice work!

Thank you!

Very nice!

Thank you Héctor!

Thank you! The robots are initialized to a rectangular formation before starting the estimation task. After this initialization step, their motion is controlled in order to collect data useful to reduce the uncertainty of their estimation.

This looks so cool that I'm pretty sure it is. Can you shortly explain how/why they moved within a rectangle pattern and why their positions soon shifted from rectangle?

nice work!