Model Predictive Control - Part 1: Introduction to MPC (Lasse Peters)
Вставка
- Опубліковано 21 лип 2024
- Introduction to Model Predictive Control; lecture presented by Lasse Peters. Recorded in Fall 2021.
#UniBonn #StachnissLab #robotics #autonomouscars #lecture - Наука та технологія
This lecture is honestly amazing, best explanation of MPC I've ever had, this will help my thesis project massively
What a timing! I needed this lecture so much.
This lecture is EXACTLY what I needed to understand MPC! Thank you so much!
This is great, thank you for taking the time to make such videos, truly great and inspiring work
What an amazing lecture. Hope that you can share more lectures about MPC. Thank you so much.
Eagerly waiting for part 2. Please do upload it asap. These lectures are amazing.
Really enjoy this video! Clearly explain the reason for using MPC before going into details.
Great lecture!! Thank you very much!!!
Great video, thank you!
Very useful and interesting lecture!
Thank you
love the video, I learnt a lot from it.
good lecture, thanks
In the example at 34:22, what is the sensor setup for localization (for describing internal state variables) of the vehicle?
Another question is that, in the robotic control at 34:57, how to get initial data for modelling the kinematics, I note that improper initialization of the model could be unstable, which cannot even make the robot running safely. This is quite different from the case of a 4-wheel vehicle, which could be safely run by a human driver to collect initial data.
thank you, this was actually very understandable.
This course is very good for understanding MPC. Is there any course related to how to implement a simulation using Python or MATLAB?
crystal clear !
at 33:50, how does the car "learn" what the track looks like? how does this get integrated into the MPC algorithm?
thank you very much
Graet lecture. can I get materials for the workshop
Hi Cyrill / Lasse, extremely helpful! Just one comment, I think that, when talking about the Model (5:25), it gives the impression that the Model element is a unique feature of the MPC concept, which I think that is not. The model is needed in standard control theory too, however I think that, in MPC, there is the possibility that the model be more complex.
yes, these kind of control problems need to know the control model or the kinematic of the robot. There are also some end-to-end deep learning models that can apply to the model-free control.
Hi, thank you for this helpful video, if I want to learn self-driving cars and from where should I start?
What about: ua-cam.com/video/EBFlmHqgezM/v-deo.html
Which my car's acceleration constraints were based on the maximal static wheel friction. Sadly mine are based on the motor :p
Hello, im korean student ,doing quadruped robot project. Can i get your PPT??
Hello, im korean uni student ,doing auto driving car project.
it's nice course! can i get your PPT??😊
Send me an email
Very nice lectures. One minor remark: The lecturer will make his lectures even more enjoyable if he looks up the pronunciation of these words (whose pronunciation he probably copied from his professor): “because”; “occur”; “estimate” (as a verb, which sounds different from when it is a noun). All the best!