Exercise 2.1f

Moderator: Lernende Roboter

Beiträge: 5
Registriert: 3. Dez 2013 15:58

Exercise 2.1f

Beitrag von roberto.calandra »

Dear Students,

As you might have noticed, the long-term prediction in this moment are "ugly". This is due to instable gains for the PD controller. Please change the gains to be:

K_p = diag([10 10])
K_D = diag([1 1])

You should now get a nice and smooth long-term prediction.

Tschuss :)

PS: also reduce the number of sampled trajectories to 20...

Zurück zu „Lernende Roboter“