search

JianZhou1212 / learning-based-rigid-tube-rmpc

This is the MATLAB code for tube robust MPC with uncertainty quantification
17 stars 8 forks source link

The pdf content of paper? #1

Open Jodorn opened 1 month ago

Jodorn commented 1 month ago

Can the author provide the pdf content of paper?

JianZhou1212 commented 1 month ago

Hi, please find the paper at https://arxiv.org/pdf/2304.05105, feel free to tell me if you have any more questions.

Jodorn commented 1 month ago

Thanks, Dr. Zhou.

JianZhou1212 commented 1 month ago

You are welcome :)

Jodorn commented 1 month ago

Dr. Zhou, I have a question. If I intend to generate one-dimensional bounded disturbance w{k} by some mechanism, can we define a conservative W according to the approximate range of one-dimensional disturbance, the actual W{true} is unknown, and then use \hat{W}^{*}{k} to approximate W{true} according to your method, so as to improve the feasibility of calculation?

JianZhou1212 commented 1 month ago

Hi, if the disturbance w_k is one dimensional, then it is very easy, you can use the same idea to learn the bound, but you can avoid solving the LP problem, just use $\hat{\mathbb{W}}^{*}_k$ = {min{\mathcal{I}_k}, max{\mathcal{I}_k}}, then you can get this learned set. I am on a conference now, and would be back around June 9, if you have follow-up questions, you can email me or Yulong (the first author).

Jodorn commented 1 month ago

Okay, thank you very much.

Jodorn commented 1 month ago

Dr. Zhou, the derivation is very complicated for me. Can you provide a detailed derivation so that I can learn it? I will be very grateful.

JianZhou1212 commented 3 weeks ago

Hi, if you do not have sufficient background in MPC, I would suggest u to start from the book: Kouvaritakis, Basil, and Mark Cannon. "Model predictive control." Switzerland: Springer International Publishing 38 (2016): 13-56. Then it would be easier for u to combine the basic things with our learning method.

Jodorn commented 2 weeks ago

OK. Thanks.