Hey everyone, thanks for watching this video! If you have any questions or comments that you'd like me to see, please leave them under this comment so that I get notified and can respond. Cheers!
Thank's Mr. #Brian for the clear explanation, I would greatly appreciate it if you kindly give us some examples about Integral action in state feedback control for disturbance rejection, thanks in advance
@@mohammedabdeldjalildjehaf1515 I don't have any plans right now to discuss integral action in an upcoming video. But I'll mention it briefly in the video I'm working on and provide some references that give good examples.
Hi Brian, For the pole placement, you could consider a butterworth configuration. Since all the poles need to be placed in state space compared to classical control, you need to find a spot for the non dominant poles. Placing them in a butterworth configuration (equidistant around a circle centered at the origin), it is shown that it approach a optimal configuration. Btw, my username on the reddit control theory sub is nickeltoes
@@marctison1039 that's a good point. I didn't really go into too much depth on where to place the poles but that would have been worth mentioning. We've spoken on Reddit, I recognized your username right away!
These videos are really good!! Great explanations. I only regret not finding them sooner. I think it filled a few gaps in my understanding and everything just suddenly clicked. If only professors gave this kind of intuitive explanations at the beginning instead of bombarding us with equations and formal definitions right away, it would make the rest of the course more interesting. Thankfully there exists the internet. Thank you Brian!
Note for college students: Please take notes and just pause and ponder at some time. It took more than a hour to completely gain some grip. Also note he might have taken a week to prepare for this set of videos. By watching just 14 mins, you just can't grab everything. If you think you are getting everything, yes definitely, but it won't last long. To remember everything he said just take notes and remember to pause and ponder. Thank you Brian for your awesome videos.
Hello @BrianBDouglas, Thank you very much for your video. I have been making some analysis on the Inverted pendulum on a cart system. Your video has allowed me to gain a deeper knowledge of how to analyze the system and what pole placement actually means. Thank you a bunch !!
Lol I'd searched Brian Douglas Pole Placement and this came first on the list. The description claimed to provide an intuitive understanding on the matter. I decided against my better judgement to click on the Matlab link as opposed to a BD one only to be greeted by the comfort of familiarity. :). Thanks Brian.
Brian, thank you! I have served myself during this quarantine with each one of your videos! You have really helped me to refresh and probably to gain a deeper understanding of control engineering. Please, keep the good work!
These series of videos contains distilled essence of many of years of experience.. Thanks to Brian and Matlab team for making these gold nuggets available free for all.. Its been tremendously helpful in my Journey towards becoming a great Control Systems Engineer!
I wish I had these in 2016 during my senior control systems class. I don't know how the prof passed me. I should have failed. I had no idea what was going on and these videos really help explain a complex topic. It's been so long that I've forgotten most of the math required to build up to this such as derivates, matrices, and eigenvalues.
My prof legit replaced himself with these video's. He just uses these video's + a book as lecture material and in the actual lectures there are just some practical examples.
Amazing video! 4:03 Matrix E for matrix A=[0 1 ; 2 -1] should be [1 -1 ; 1 2]. What am I missing? Where do the values in this matrix E shown in the video come from?!
Hello, Brain. Firstly, thank you for your continuous sharing. And then, I want to ask you a question. In this video at 8:10, you said" The gain matrix K is a 1 by 2 since there's one OUTPUT and two states." But as you said in the previous slide, kx is part of u. And u is the input. So, why isn't the gain matrix K is a 1 by 2 since there's one INPUT and two states?
That is some excellent sleuthing skills! I did mean to say input there. I'll see if there is a way MATLAB can add a card or a note in the description to call out this error. Thanks for finding that and letting me know!
your state space diagram looks differently to the one shown in the book by powell and franklin. How can i reconcile these differences? Thanks for the video :)
How do you calculate the reference gain when there are multiple inputs/outputs? For instance, if the system is 2x2, then the DC gain will be 2x2, do you invert the matrix and multiply by the reference?
Hi Brian, Thanks for all your great videos. I have a question regarding the control signal u for the PD controller at 12:36. Shouldn't the control signal u be a function of the error (e = r-p) hence the controller output u should be: u = Kp.e + Kd.e_dot = Kp.(r-p) + Kd.(r-p)_dot ? I mean shouldn't r be multiplied by the gains also?
@@BrianBDouglasMaybe you meant a special case. For example if the input r is constant (a step input) , then its derivative will be zero. So we'll end up with r.Kp which makes it similar to the state feedback equation with Kr=Kp and the rest are the same.
@@BrianBDouglas why isn't B multiplied by Kr as well when calculating the modified A matrix? i.e. Acl = A - B*Kr*K? Asking because B is multiplied by Kr later.
Hello Brian, your video is amazing. Could you make video on how we deal with N order of system in terms of controller design and analysis , i remember you mention one of your video you have worked on aerospace industry and you know aircraft control has more than 10000 order of model equation need to be aken care of if you could explain the strategy for them it would be helpful and if you could make a video on MPC specially explaining the mathematics or core algorithm behind it, it would be helpful.thnx.
You can use root-locus design method techniques to find the locations that meet your requirements. Usually when working in a project you have some requirements regarding damping (or overshoot), bandwidth, loop gain and actuator effort (or something related to actuator effort, like battery life of total system).
Great video! Question though, How would I use MATLAB to solve for the poles if I am given the natural frequency and damping ratio? I'm given: u(t) = -K( x - x_d) My matrices A-D the size of K is a 1x4 vector any tips would help thanks!
Brain, another great video as usual 👍. I especially like your discussion on practical implementation issues at 13:04. If you're interested, I actually made a video discussing some of these issues at ruclips.net/video/9vCTokJ5RQ8/видео.html. I'm looking forward to more of your videos!
Hey everyone, thanks for watching this video! If you have any questions or comments that you'd like me to see, please leave them under this comment so that I get notified and can respond. Cheers!
Thank's Mr. #Brian for the clear explanation,
I would greatly appreciate it if you kindly give us some examples about Integral action in state feedback control for disturbance rejection,
thanks in advance
@@mohammedabdeldjalildjehaf1515 I don't have any plans right now to discuss integral action in an upcoming video. But I'll mention it briefly in the video I'm working on and provide some references that give good examples.
how do I select the values in the reference input 'r', if I only care about one state (lets say, speed) and not the others (lets say, current)?
Hi Brian,
For the pole placement, you could consider a butterworth configuration. Since all the poles need to be placed in state space compared to classical control, you need to find a spot for the non dominant poles. Placing them in a butterworth configuration (equidistant around a circle centered at the origin), it is shown that it approach a optimal configuration.
Btw, my username on the reddit control theory sub is nickeltoes
@@marctison1039 that's a good point. I didn't really go into too much depth on where to place the poles but that would have been worth mentioning. We've spoken on Reddit, I recognized your username right away!
These videos are really good!! Great explanations. I only regret not finding them sooner. I think it filled a few gaps in my understanding and everything just suddenly clicked. If only professors gave this kind of intuitive explanations at the beginning instead of bombarding us with equations and formal definitions right away, it would make the rest of the course more interesting. Thankfully there exists the internet. Thank you Brian!
100% agreed
so true!
Man, you are awesome. I understood in 14 minutes what i couldn't in a semester. Thanks a lot!
yes
I second this
Brian has been always the best in Control field since i was learning a lot from his channel. Thanks Brian
Brian is definitely the best. Glad to see you are working with MATLAB now. I still visit your videos to ground myself in control foundations. Kudos!
Note for college students:
Please take notes and just pause and ponder at some time. It took more than a hour to completely gain some grip.
Also note he might have taken a week to prepare for this set of videos. By watching just 14 mins, you just can't grab everything. If you think you are getting everything, yes definitely, but it won't last long. To remember everything he said just take notes and remember to pause and ponder.
Thank you Brian for your awesome videos.
Hello @BrianBDouglas, Thank you very much for your video. I have been making some analysis on the Inverted pendulum on a cart system. Your video has allowed me to gain a deeper knowledge of how to analyze the system and what pole placement actually means. Thank you a bunch !!
Dude, u’ve helped me in my undergrad, now you are helping me with masters. Thanks
same here.
just wow, I came from a class where they teach us all of these in a different way, but checking the overview made it much clear
Thank you sir. As you said in the video: you are the lightning making the world of contol theory so bright!
I love it. I understood very quickly what I was struggling to get in my class!!
Great!
This was wildly satisfying to watch after taking my controls classes.
Lol I'd searched Brian Douglas Pole Placement and this came first on the list. The description claimed to provide an intuitive understanding on the matter. I decided against my better judgement to click on the Matlab link as opposed to a BD one only to be greeted by the comfort of familiarity. :). Thanks Brian.
Brian, thank you! I have served myself during this quarantine with each one of your videos! You have really helped me to refresh and probably to gain a deeper understanding of control engineering. Please, keep the good work!
These series of videos contains distilled essence of many of years of experience.. Thanks to Brian and Matlab team for making these gold nuggets available free for all.. Its been tremendously helpful in my Journey towards becoming a great Control Systems Engineer!
I really regret not finding these series before >>> Now control theory starts making sense for me
I'm happy to hear that this video helped ☺
I wish I had these in 2016 during my senior control systems class. I don't know how the prof passed me. I should have failed. I had no idea what was going on and these videos really help explain a complex topic. It's been so long that I've forgotten most of the math required to build up to this such as derivates, matrices, and eigenvalues.
this video has changed my life
please replace my professor.
*our professors
can't like this comment, can't ruin the perfect number of 119 likes
My prof legit replaced himself with these video's. He just uses these video's + a book as lecture material and in the actual lectures there are just some practical examples.
He doesn't even have to replace him, just get rid of him ....
Agree 💯 @@abdelrahmanmahdy9354
Amazing video!
4:03 Matrix E for matrix A=[0 1 ; 2 -1] should be [1 -1 ; 1 2]. What am I missing? Where do the values in this matrix E shown in the video come from?!
Genius teaching ! Really awesome ! From a french student
Such a great work, I really hoped the video wouldn't end. Thanks!
amazing. I'm just amazed by a fact that some genius figured this out a long time ago. How did he or she come up with it!? haha
giants sitting on the shoulders of other giants, all the way back to pythagoras
man, this video deserves more likes
YOU JUST SAVED MY LIFE
These videos are seriously underrated. Great work!
Great job with these videos Brian! Love how you add color to these equations...makes it so much easier to follow along.
oh man, you are the best of the best teacher. brief and useful
Thanks
that's great
we need videos about the nonlinear systems such the "negative resistance oscillator"
we will be very thankful
inspring explanation and animation for mordernized control theroy, thanks a lot for shaing!
Lovely explanation in a short time thanks MATLAB
Wow! This video explains so many concepts so well. Thank you Brian!
Excellent explanation
Awesome, clear and direct to the point as usual. Love you Bud.
You are my outstanding teacher.
This video was pure 🔥🔥🔥
Very good explanation and examples, thanks.
This video was so well done! Bravo Brian, Bravo!
Another 👍 Brian's video... Thank you from my ❤
Hi Brian , Thanks for providing a beautiful lecture. could you please share an example with real system like a DC motor servo.
saneemp, if you're interested here is an example of full state feedback with a DC motor system ruclips.net/video/9vCTokJ5RQ8/видео.html
Hello, Brain. Firstly, thank you for your continuous sharing. And then, I want to ask you a question. In this video at 8:10, you said" The gain matrix K is a 1 by 2 since there's one OUTPUT and two states." But as you said in the previous slide, kx is part of u. And u is the input. So, why isn't the gain matrix K is a 1 by 2 since there's one INPUT and two states?
That is some excellent sleuthing skills! I did mean to say input there. I'll see if there is a way MATLAB can add a card or a note in the description to call out this error. Thanks for finding that and letting me know!
@@BrianBDouglas HAHAHA. It's my pleasure. It feels good to hear from you.
your state space diagram looks differently to the one shown in the book by powell and franklin. How can i reconcile these differences? Thanks for the video :)
Your explanations are amazing
Amazing video. You make my weekend better!
How do you calculate the reference gain when there are multiple inputs/outputs?
For instance, if the system is 2x2, then the DC gain will be 2x2, do you invert the matrix and multiply by the reference?
you're videos are the best, this really saved me on my final. Thanks so much!
Hi Brian, Thanks for all your great videos.
I have a question regarding the control signal u for the PD controller at 12:36.
Shouldn't the control signal u be a function of the error (e = r-p) hence the controller output u should be: u = Kp.e + Kd.e_dot = Kp.(r-p) + Kd.(r-p)_dot ? I mean shouldn't r be multiplied by the gains also?
Gah! I completely messed that up. I'm not sure what I was thinking when I wrote that transfer function out. Good catch!
@@BrianBDouglasMaybe you meant a special case. For example if the input r is constant (a step input) , then its derivative will be zero. So we'll end up with r.Kp which makes it similar to the state feedback equation with Kr=Kp and the rest are the same.
Thanks for asking the question. I thought I was going wrong somewhere.
@@BrianBDouglas why isn't B multiplied by Kr as well when calculating the modified A matrix? i.e. Acl = A - B*Kr*K? Asking because B is multiplied by Kr later.
i also has this doubt thanks from your reply I got ans
10:15 Why use B matrix as ss closed-loop second argument?
Beautiful.
Only, I haven't been able to find a good definition for the poles. Can you direct me to one?
Brian you are great you explain so we'll 😍😍
Thank you so much! This was really helpful.
Hi Brian, Good work!, will you tell us about the linearization of non linear model? thanks
In this approach , we can't look in the system to be adjusted by phase margin like in scalar approach.
😁😁😁clear, concise, brilliant explanation
Great explanation!!! Thanks a lot...
why do i get for E by min 4:10 for [V,D] = eig(A) in MATLAB for V column 1 and 2 switched? [0.71,-0.45; 0.71, 0.89]
Can anyone explain the foll. please:
[1 0] *[k1 k2] = k1 but in the example there is k2 as well
Hello Brian, your video is amazing. Could you make video on how we deal with N order of system in terms of controller design and analysis , i remember you mention one of your video you have worked on aerospace industry and you know aircraft control has more than 10000 order of model equation need to be aken care of if you could explain the strategy for them it would be helpful and if you could make a video on MPC specially explaining the mathematics or core algorithm behind it, it would be helpful.thnx.
Can someone explain to me how we get the E matrix for the modal form?
Anyone ?
Wonderful .. Great Lec.
Bravo 🤝
This video is amazing!
Thanks a lot for wonderful video.
Hello, X’=AX+BU, let V=AX+BU then U=inv(B)(V-AX) => X’=V and Could we choose V=-KX and U=-inv(B)(K+A)X?
Thankyou #Brian Sir.
Well Done! Thanks!
Is the control input equation for PD controller at 12.35 correct?
Thanks
why is it showing error while using poles imaginary ??
Thank you it was awsome
AWESOME !!
very nice thanks
is there a reason why we move the pole values to -1 and -2 or just randomly choose, is there a method to choose the best poles for the system
You can use root-locus design method techniques to find the locations that meet your requirements.
Usually when working in a project you have some requirements regarding damping (or overshoot), bandwidth, loop gain and actuator effort (or something related to actuator effort, like battery life of total system).
thank you soooo much
thanks bro
respected Matlab, is it possible to do hacking wid help of Matlab?
i feel proud since i am an electrical and computer engineering (power)student
THANKS ! S2
Great video! Question though, How would I use MATLAB to solve for the poles if I am given the natural frequency and damping ratio?
I'm given:
u(t) = -K( x - x_d)
My matrices A-D
the size of K is a 1x4 vector
any tips would help thanks!
Hello Brian Douglas
All superheroes dont wear capes .
Best
Brain, another great video as usual 👍. I especially like your discussion on practical implementation issues at 13:04. If you're interested, I actually made a video discussing some of these issues at ruclips.net/video/9vCTokJ5RQ8/видео.html. I'm looking forward to more of your videos!
Whose here during Covid trying to substitute their university shitty online learning?
Almost everyone