Day 2 - Development of 3 PD controllers

These last days I’ve been working on the development of the exercise of the subject “Robotic Vision”, in particular, I have successfully developed 3 PD controllers that manage to finish the lap of the circuit “Default Line” in a time between 1’ 35’’ and 1’ 45’’.

To achieve these results, I had to invest a lot of time in orchestrating the three controllers, because one of them is for very steep curves, another one for lighter curves, and another one for straights. In addition, once I decided when to use each controller, I had to adjust the K_p and K_d constants (proportional constant and derivative constant) very well.

The ecuation for these PD controllers is:

\[u = Kp*e + Kd * de/dt\]

On the other hand, I have also worked on making a more robust perception than the one I had in the previous post. Specifically, before calculating the midpoint of the red line at a given height, I applied 3 dilations and 3 erosions on the resulting image of the color filter, in order to eliminate noise.

Here you can watch a video of the result:

In the next post, I will implement 3 PID controllers and compare the results with this ones. With that, my exercise will be finished.