Keep going and don't stop uploading your videos. I am an Asst. Professor and I explain to my students after watching videos all over the internet. I think you explain very well and also you have good examples in your videos! Simple and Effective...Cheers from India
Man, your videos are gold. Control theory gold. I wish I had seen this when I was at the university! you definitively go deep in a way people actually grasp the concepts and ideas! keep with the excelent videos!
Spotted a typo: at 9:52 in the video, the Matlab comment text says "But a downside to rlocus is that you can't see how all of the roots move at the same time, you only have control over one root. Also, you can easily see the effect of" The second "can" should be a "can't". Anyway, this is an excellent video series, Brian Douglas!
Great demonstration and walkthrough. I had not used sisotool before, what an eye opener! If you have more matlab walkthroughs like this with other tips and tricks for e.g. Bode plot that would be greatly appreciated. Keep up the good work
Thank you so much for your effort and dedication in making these videos! They are really awesome and helped me pretty much to understand a lot of topics of classic control. When you have time, could you make some videos talking about state space control theory? Thanks in advance! Best Regards!
@9:25 we can see that the popup shows a damping ratio term that is less than one. The question is, The damping ratio is defined only for the second order system (as stated earlier in the video). How are we applying the same concept here in the fourth order system ?
first of all thanks a lot Brian !! This is one of the best videos which I have watched from your Control System Lectures videos and If I say in terms of MATLAB for Control Engineering than probably this would be one the best videos of MATLAB. I am Instrumentation and Control Engineering student but I had never known this features of MATLAB before. This video will be very helpful for learning purpose but Brian I request you to make some videos on "Modelling of DC Motor with 'encoder' or 'Accelerometer and Gyroscope' using Arduino and MATLAB/LabVIEW" as well as on "How to Implement PID algorithm in Arduino for DC motors for the Motion Control of Robots". So that there won't be only learning of these features of MATLAB but also practical hands on and understanding of how to use these tools in Robotics and other applications. Eagerly waiting for those videos... and once again, thanks for this awesome video. cheers !!
terrific! i would like to know something maybe trivial when you design a controller, you are adding notch, zeros...in order to meet the requirements. but can you add whatever you want?i mean what is the meaning in term of hardware, what is the cost of this?thank you a lot
Tony Sznabel Not trivial at all! When you're working in the S-domain and just adding poles and zeros it's easy to lose track of what that actually means. Whatever controller you design you have to build - and more complicated designs require more components. But beyond that, you might actually build a controller that attempts to drive your actuators faster than they can be driven and that's not going to work. For example, I can design something that controls a motor so that it has a step response rise time of 1 millisecond and my simulation will say no problem. But if the system can't source that much current, or that high of voltage (or if the change in current doesn't occur fast enough) then the motor actually won't accelerate as fast as my controller is driving it. The best way to learn all of this is to take on a small project at home and see how your real hardware behaves compared to the ideal model you've assumed. Hope that helped a bit.
Awesome videos ..Loved it .... Can you make video with some real life problem and solving it using Matlab ? (Tools used can be root locus , Bode Plot ,Nyquist PLot ,Lead and Lag Compensator ,PID Tune ) That would be an amazing experience to learn under you.Thank you:)
The videos are very helpful in order to understand the deep concepts behind the Root Locus technique. But I do have a pretty confusing question, actually two questions. I have understood all the concepts but this fundamental one is confusing me a bit. My first question is that we use Root Locus in order to analyze our closed loop system like how the poles of our closed loop system move as we vary ''k". So then why do we plot the root locus of GH only why not 1 + GH? My second question is pretty related to the first one. In one of your video, you said that our closed loop poles travel from the open loop poles to the open loop zeros. What does that mean?
+Nan Nanco You are right that the characteristic equation is the whole denominator. I wasn't very clear but what I meant was that we needed the second part of the characteristic equation in the form that it's in, that is specifically, in the G*H form. And since the problem gave us the open loop transfer function as G*H that was exactly what we needed. So the characteristic equation was in the form 1 + k*G*H that we needed to plot the root locus. Hope that helped a bit.
So in fact, we should draw the root locals for 1+kGH (closed loop) and not GH(open loop). Anyway, the question ask to draw for GH, so we just draw it. (I think this is what Brian meant?)
@@Ming_Qiu When K is equal to zero, the limit of the denominator approaches the denominator of G*H, which means that the closed-loop poles approach the open-loop poles. When K approaches infinity, the limit of the denominator approaches the numerator of G*H. The root locus tracks the path of the solutions to the equation of 1 + K*G*H = 0, on K's journey from 0 to infinity, such that the closed loop poles follow a path from the open loop poles to the open loop zeros. That's the essential idea of the Root Locus
Hi Brian! First of all, thank you for helping us to understand more about this topic. I have a question. So when we use the SISO tool, and we take a look at the RootLocus, i noticed that the 'Squares' are not in the same position as the poles (for a k=1), with this transfer function you have to zoom in a lot, but i've noticed that with others transfer functions, this is more visible. So at the end using a k=1 where are our poles? Where the squares are or where the X's are?
This video series is exactly what I'm looking for. Btw I have a question regarding the stack TF mentioned above. I'm working with a ship-heeling-angle control system in my uni lab. The system includes two tanks located on two sides of the hull and a gear pump which pumps water back and forth into the tanks depending on the input voltage negative or positive. After conducting several experiments to determine the step responses with various step reference voltages, I figure out that the transfer functions might not look the same. I haven't done the system identification step yet, however from examining the graphs (which look like an integrating system with a time delay element) it seems like the the gain constant coef K differ from eachother. I wonder if I should work with these transfer function as a stack or you know choose one that represents the whole system? This question hasn't popped up in my head until I watched this interesting video.
Just to leave a feedback from a not-student-anymore, or rather, an once-upon-a-time-a-student: concluded my studies on control systems in 2000, but not professionally on this field since.. but the videos are being extremely useful for me to dust off these knowledges deep buried in 15 years, and more importantly, putting some intuitive light on the dull, hard, at the time, abstract mathematics, to put some logic and understanding on the behaviour of these systems and methods. Sorry for not putting any interesting questions at this time :)
awesome, can you help with how can i introduce any transfer function, i mean i want to make a programm which could ask for any transfer function and it could plot rlocus method
I have a control system where there is a disturbance that only comes into effect after 100 seconds. I am wondering how I incorperate this into the transfer function of the system? Any help would be greatly appreciated
Why, if the gain increases as you move through the root locus, is the step response stabilizing always at 1? Shouldn't the unitary step be multiplied by the gain in question?
Question: Write a code to simulate the root locus for any simple transfer function G(S) H(S) and verify the solution by sketching the root locus for the same transfer function choosen.
Man this makes me appreciate the people behind this package. They're some smart people. Also thank you for making these videos!
Keep going and don't stop uploading your videos. I am an Asst. Professor and I explain to my students after watching videos all over the internet. I think you explain very well and also you have good examples in your videos! Simple and Effective...Cheers from India
You explained how to use the root locus plots on MATLAB 100x better than all my professors. Thank you!
Man, your videos are gold. Control theory gold. I wish I had seen this when I was at the university! you definitively go deep in a way people actually grasp the concepts and ideas! keep with the excelent videos!
i wanna ask you bro,how the aplication of this teory?i just know the teory
Spotted a typo: at 9:52 in the video, the Matlab comment text says "But a downside to rlocus is that you can't see how all of the roots move at the same time, you only have control over one root. Also, you can easily see the effect of"
The second "can" should be a "can't".
Anyway, this is an excellent video series, Brian Douglas!
I really appreciate your videos.
Manuel Cortez Thanks!
I like the efficiency you convey information in a short time in a clear and concise manner! Definitely will watch several times.
Incredible feature of sisotool that you show in the last part this video. That is a useful help for all. Thanks!
Thank you my guy, you just saved my grade. It's a holiday miracle.
man you are extremely good at teaching & your matlab understandig is absolutely amazing ! congrats!! keep on teachig you are doing great !!
The Sisotool is awesome. I didn't know about it until now. What it can do is just amazing and a lot! Quite helpful
this is the best channel on the planet, thank you so much
HI, I'm from Brazil, and I really appreciate your lectures! Simple awesome! Congratulations!
I really hope you could become a lecturer in my university . I wouldnt have skipped so many classes if you were my teacher
Great demonstration and walkthrough. I had not used sisotool before, what an eye opener! If you have more matlab walkthroughs like this with other tips and tricks for e.g. Bode plot that would be greatly appreciated. Keep up the good work
Man! I think you're completely MASTER!
You are going to heaven for these videos
I've watched a lot of your videos and learned a ton. This one was something special! Thanks a lot!
Thanks Brian
You're going straight to the point
Your explanation is very helpful
Thank you so much for your effort and dedication in making these videos! They are really awesome and helped me pretty much to understand a lot of topics of classic control. When you have time, could you make some videos talking about state space control theory? Thanks in advance! Best Regards!
Really awesome video!!!! Helped me out SO MUCH you have no idea! Thanks a lot! I hope you find great success on TH-cam!
Excellent video, Brian! This tutorial is really effective and simple to follow! Thank you for your time.
@9:25 we can see that the popup shows a damping ratio term that is less than one. The question is, The damping ratio is defined only for the second order system (as stated earlier in the video). How are we applying the same concept here in the fourth order system ?
Thank you very much for posting those videos and I really appreciate .
"Gain" a better understanding of Root Locus plots? Was that a pun?
CanadianRocketry Hah, no not intentional. Would you prefer "Get to the root of the problem with Matlab"?
or a complete Matlab tut on plotting poles and zeros
Nice one
Incredible videos! Cleared out many questions I had in mind.
This was the coolest thing i hv ever seen
Really good video, great help when studying root locus.
ขอบคุณมาก เป็นการสอนพื้นฐานที่เข้าใจมากครับ Thank you very much for making this video.
very helpful tuto... I appreciated it enough .. only explaining the essential
Very useful , your knowledge on matlab is also great. Thank you very much for your videos !
Don't stop making videos like this ...
Thanks for uploading this one.
Simply impeccable.
🔥🔥🔥 so cool, u make this course so interesting when compared to my prof
Super thumbs up, this was super useful. Thank you Brian.
Thank you for your efforts, great vids
Keep up the awesome work!
YOU ARE THE SYSTEM CONTROL GOD!!
Your videos is more than helpful ..Thanks
Continue... :)
thank you for making this video i really benefit a lot
This video is better than the new Star Wars.
Most seems to work in MATLAB 2024 as well.
Great video.
first of all thanks a lot Brian !!
This is one of the best videos which I have watched from your Control System Lectures videos and If I say in terms of MATLAB for Control Engineering than probably this would be one the best videos of MATLAB.
I am Instrumentation and Control Engineering student but I had never known this features of MATLAB before.
This video will be very helpful for learning purpose but Brian I request you to make some videos on "Modelling of DC Motor with 'encoder' or 'Accelerometer and Gyroscope' using Arduino and MATLAB/LabVIEW" as well as on "How to Implement PID algorithm in Arduino for DC motors for the Motion Control of Robots". So that there won't be only learning of these features of MATLAB but also practical hands on and understanding of how to use these tools in Robotics and other applications.
Eagerly waiting for those videos...
and once again, thanks for this awesome video.
cheers !!
you are legend man wish to u all the best and thank u so much
Woa this is very cool, please make more control system MATLAB video.
great explaination of matlab for root locus
This was really good!
You are awesome!
Thank you for doing this!
thanks sir your videos are really awsome.plz sir post something on block diagram questions.. it seems complicated solving them .
You are the best man!
terrific!
i would like to know something maybe trivial
when you design a controller, you are adding notch, zeros...in order to meet the requirements. but can you add whatever you want?i mean what is the meaning in term of hardware, what is the cost of this?thank you a lot
Tony Sznabel Not trivial at all! When you're working in the S-domain and just adding poles and zeros it's easy to lose track of what that actually means. Whatever controller you design you have to build - and more complicated designs require more components. But beyond that, you might actually build a controller that attempts to drive your actuators faster than they can be driven and that's not going to work. For example, I can design something that controls a motor so that it has a step response rise time of 1 millisecond and my simulation will say no problem. But if the system can't source that much current, or that high of voltage (or if the change in current doesn't occur fast enough) then the motor actually won't accelerate as fast as my controller is driving it. The best way to learn all of this is to take on a small project at home and see how your real hardware behaves compared to the ideal model you've assumed. Hope that helped a bit.
Thank You very much Your videos are very good
This is really fantastic because you re fantastic
Thanks man, this video very useful.
has the sisotool changed because mine looks completely different and I can't do this stuff anymore?
Great tutorial bro
tahnk you Brain this was very intresting
Thank you so much...This video very useful.
NICE!!! Thank you! Great video.
Brian Brilliant !! Another great video.
Could you pls talk about the impact of the "pure delay" in the stability of the system.
bob jorges Pure delay and Z-transform is a great video topic. Added to the list!
Awesome videos ..Loved it .... Can you make video with some real life problem and solving it using Matlab ? (Tools used can be root locus , Bode Plot ,Nyquist PLot ,Lead and Lag Compensator ,PID Tune ) That would be an amazing experience to learn under you.Thank you:)
lo amo señor , muchas gracias
At 7:00 you say that you are going to plot the pzmap for gains of k from 0 to 10, but you actually loop through k from 1 to 10
+Anthony Zhu So it is the same thing, in fact going through the loop with the values of k from 0-10 is good stuff rather than plotting them one by one
The videos are very helpful in order to understand the deep concepts behind the Root Locus technique. But I do have a pretty confusing question, actually two questions. I have understood all the concepts but this fundamental one is confusing me a bit.
My first question is that we use Root Locus in order to analyze our closed loop system like how the poles of our closed loop system move as we vary ''k". So then why do we plot the root locus of GH only why not 1 + GH?
My second question is pretty related to the first one. In one of your video, you said that our closed loop poles travel from the open loop poles to the open loop zeros. What does that mean?
1:35 Why did we omit the "1+" in the closed loop TF? Isn't the WHOLE denominator Characteristic Equation ?
+Nan Nanco You are right that the characteristic equation is the whole denominator. I wasn't very clear but what I meant was that we needed the second part of the characteristic equation in the form that it's in, that is specifically, in the G*H form. And since the problem gave us the open loop transfer function as G*H that was exactly what we needed. So the characteristic equation was in the form 1 + k*G*H that we needed to plot the root locus. Hope that helped a bit.
+Brian Douglas I think i got it now, thanks for the reply and for all the videos!
So in fact, we should draw the root locals for 1+kGH (closed loop) and not GH(open loop). Anyway, the question ask to draw for GH, so we just draw it. (I think this is what Brian meant?)
@@Ming_Qiu When K is equal to zero, the limit of the denominator approaches the denominator of G*H, which means that the closed-loop poles approach the open-loop poles. When K approaches infinity, the limit of the denominator approaches the numerator of G*H.
The root locus tracks the path of the solutions to the equation of 1 + K*G*H = 0, on K's journey from 0 to infinity, such that the closed loop poles follow a path from the open loop poles to the open loop zeros. That's the essential idea of the Root Locus
Hi Brian! First of all, thank you for helping us to understand more about this topic. I have a question. So when we use the SISO tool, and we take a look at the RootLocus, i noticed that the 'Squares' are not in the same position as the poles (for a k=1), with this transfer function you have to zoom in a lot, but i've noticed that with others transfer functions, this is more visible. So at the end using a k=1 where are our poles? Where the squares are or where the X's are?
This video series is exactly what I'm looking for.
Btw I have a question regarding the stack TF mentioned above.
I'm working with a ship-heeling-angle control system in my uni lab. The system includes two tanks located on two sides of the hull and a gear pump which pumps water back and forth into the tanks depending on the input voltage negative or positive.
After conducting several experiments to determine the step responses with various step reference voltages, I figure out that the transfer functions might not look the same. I haven't done the system identification step yet, however from examining the graphs (which look like an integrating system with a time delay element) it seems like the the gain constant coef K differ from eachother.
I wonder if I should work with these transfer function as a stack or you know choose one that represents the whole system?
This question hasn't popped up in my head until I watched this interesting video.
helped a lot. many thanks
Just to leave a feedback from a not-student-anymore, or rather, an once-upon-a-time-a-student: concluded my studies on control systems in 2000, but not professionally on this field since.. but the videos are being extremely useful for me to dust off these knowledges deep buried in 15 years, and more importantly, putting some intuitive light on the dull, hard, at the time, abstract mathematics, to put some logic and understanding on the behaviour of these systems and methods. Sorry for not putting any interesting questions at this time :)
I can't even understand what the hell you said.
Dude thanks a million.
Awesome man , really thanks
Would you recommend MATHEMATICA and SYSTEMMODELER for Control System Analysis ? Or MATLAB would be a better choice for this purpose ?
im using matlab 2020a windows 10, only location, damping and natural f shown at the bottom, any idea how can i get the gain value?
you saved hundreds of reading hours in reading textbook
you are an awesome teacher, thanks for all your videos. can you do a video on predictive control (e.g. smith predictor)?
nice work....... u r great bro
awesome, can you help with how can i introduce any transfer function, i mean i want to make a programm which could ask for any transfer function and it could plot rlocus method
Neat video. What's a nichols diagram that looks like a spider ?
I have a control system where there is a disturbance that only comes into effect after 100 seconds. I am wondering how I incorperate this into the transfer function of the system? Any help would be greatly appreciated
did a great job
Why isn't the transfer function at the beginning of the problem in the form of Gcl = (kGH/(1+kGH) ?
Just messing around with the pzmap function for a second order PI controller system, how do you get the grid lines and zeta lines to map in the plot
Type "grid on" in the command window
you r really good, thanks a lot
Great lessons. Do you have any plans in presenting an equivalent analysis in the Z domain?
My muscle "Gains" have been dropping recently. Should i use a lead or lag compensator?
Great videos, could you please share a video on Digital PI controller for a motor current loop.
Very useful!
How can I draw root locus for system with PR controler [1+Ki*s/(s^2+50^2)] and the object is G = (s+1)/(s^3+10s+15) when Ki change
What if K is on the bottom, how do you variate it so you can see the differences
Why, if the gain increases as you move through the root locus, is the step response stabilizing always at 1? Shouldn't the unitary step be multiplied by the gain in question?
Because that step response is for the closed loop system. The plot of "step(GH*k)" and "step(feedback(GH*k,1))" are different.
@@tutimoyano Thanks!
Thanks very helpful!
Question:
Write a code to simulate the root locus for any simple transfer function G(S) H(S) and verify the solution by sketching the root locus for the same transfer function choosen.
how did u get k g(s) h(s)....at the beginning of the vid
How to do the stability analysis using root locus in matlab? Please help me
what version of matlab do you use?
Which software is used for writing?
Amazing thank u a lot. .
awesome man!!!!!
I dont know about you guys but when I open control system designer I only get one window the design window and I can't find a manager window anywhere.
how to get k value in transfer function brotha?