A PID Tuning Guide | Understanding PID Control, Part 4

Поделиться
HTML-код
  • Опубликовано: 23 дек 2024

Комментарии • 78

  • @BrianBDouglas
    @BrianBDouglas 6 лет назад +169

    Hey everyone, thanks for watching! I know this was a whirlwind of a video so if you have any questions on any of it leave them here and I'll try my best to answer them. Cheers!

    • @vasylvarvolik4332
      @vasylvarvolik4332 6 лет назад +3

      Hi Brian)
      It is very interesting. You explained to me very clear.

    • @BrianBDouglas
      @BrianBDouglas 6 лет назад +2

      Thanks for the feedback!

    • @ehsanu1
      @ehsanu1 6 лет назад +3

      Great video! I wasn't 100% sure what you meant by an integrator-like plant, despite the example of a motor. Is that idea that a motor, given a constant signal, over time accumulates distance in whatever it's driving? It's not clear to me how that would obviate the need for an integral factor in the controller.

    • @ehsanu1
      @ehsanu1 6 лет назад +2

      Oh, and can you point to any resources to help with the "understand the system you're controlling" bit?

    • @BrianBDouglas
      @BrianBDouglas 6 лет назад +8

      That was confusing, wasn't it? A motor in position control acts like an integrator for the reason you state - it accumulates distance given a constant signal. The motor control example I used was a velocity controller, which doesn't act like an integrator since a constant voltage produces a constant velocity. Therefore, my statement about the motor controller was misleading. To understand why an integrating transfer function may not require an integrator in the PID controller, search Google for "transfer function type steady state error" and click on the UMich link at the top. There is a great explanation of transfer function type and how that relates to steady state error. If you want zero steady state error for a step input then your system needs to be a type 1 (or one pole at the origin, or another way of putting it - an integrator). If your plant doesn't have that integrator, then you'll need to add one in the controller. Hope this helps.

  • @DeAdMaN148
    @DeAdMaN148 12 дней назад +1

    I LOVE YOU BRIAN DOUGLAS! You've made a tremendous series. Thank you for contribution to humanity.

    • @MATLAB
      @MATLAB  11 дней назад

      I'm glad you enjoyed the series! Brian is awesome!

  • @videoexpert27
    @videoexpert27 3 года назад +13

    This was a nice overview of what I have learned in university. In the courses the things didn't form a good big picture but this helps me to put everything together so the knowledge becomes more useful

  • @tbk01
    @tbk01 4 года назад +8

    You make stuff so much more understandable. the PID tuning flowchart was incredible!

    • @mnada72
      @mnada72 3 года назад +3

      Where is it ?

    • @Deuto418
      @Deuto418 3 года назад +2

      @@mnada72 Link in the description after: PID Control System Analysis and Design - IEEE Article. It is the first table in the linked article. (edit: The Flow Chart is in the video, for example at 11:30)

  • @seenathgooding181
    @seenathgooding181 4 года назад +3

    Great Video. This series is giving me an understanding that my lecturer couldn't. Thank you.

  • @bharathp1000
    @bharathp1000 6 лет назад +2

    Hi Brian..I guessed it was you , that presentation and voice won't (your channel) forget that.,,
    Best explanation.👍

  • @quinn6152730
    @quinn6152730 4 года назад

    I didn't click the like bottom every part but I have to say every part is crystal clear and the order is well arranged!! Thank you matlab for saving a disappointed ME student doing project^^

  • @fifaham
    @fifaham 2 года назад +2

    I once worked in a Florida based company that manufactres security products and I implemented an algorithm (C and Assembly code) in the microcontroller that perfoms "Automatic-Self-Testing" for the product to save time testing the product after production, this also saved time and money on testing the product. The reason I metioned this is that now I am thinking of similar automatic-self-tuning algorithm that will drive the Drone to the desired behaviour and automatically adjust the PID gains over time to compensate for the aging and deformation of Dron parts and sensors. I think this kind of algorithm will solve the tuning problem of PID control loop. Has anyone done anything like that before? I guess such algorith may need a high MIPS ARM microcontroller with very generous memory. Once this is a success I will add it to the service of my small business.

    • @sandmaster4444
      @sandmaster4444 Год назад +1

      Maybe a bit late, but for anyone interested in this kind of work, look into adaptive control

  • @rexsolomon6325
    @rexsolomon6325 4 года назад +1

    Hello Brian, your videos are fantastic! A perfect tie-in for this Part4 PID Tuning would be Benjamin Vedder's VESC Tool, specifically, the PID Controller menu tab (the default numbers are meant to be tweaked for different motors, but there is no clear guide as to how). If you could integrate Benji's VESC Tool PID Controllers into your Part 4 PID Tuning video (as a new video), it would be of tremendous, immediate benefit to thousands of e-bike, e-skateboard and quadcopter users. Thank you!

  • @TheSymonak
    @TheSymonak 5 лет назад +3

    What's "table 1" to see the gains on PID effects?

  • @blessoneasovarghese9834
    @blessoneasovarghese9834 5 лет назад +11

    Where is the handy chart of PID tuning for different applications?

    • @ahnafsakib
      @ahnafsakib 5 лет назад

      It's in table 1 of this paper

    • @mnada72
      @mnada72 3 года назад

      @@ahnafsakib Where is it ?

  • @SMV1972
    @SMV1972 Год назад

    Брайан, спасибо большое за знания!

  • @СтасиаЕвдокимова-л6ч
    @СтасиаЕвдокимова-л6ч 2 года назад +1

    Brian ! You are amazing ,amazing!🙌😱🙏

  • @abhinavbharadwaj9755
    @abhinavbharadwaj9755 10 месяцев назад

    PID tuning has never been explained in a more appreciable form

  • @renatoduenas4104
    @renatoduenas4104 5 лет назад

    Awesome I am checking all your videos. Your method to explain is great. Congratulations!

  • @leadeeeeer
    @leadeeeeer 6 лет назад +2

    Brian, This is amazing! thanks so much !

  • @scotthebert-d2k
    @scotthebert-d2k Год назад

    what is 'table 1' in the description supposed to point us to? I was looking forward to the gains guide

    • @HansScharler
      @HansScharler Год назад

      That is a funny reference. Here's the table: blogs.mathworks.com/community/files/table-1.png

    • @BrianBDouglas
      @BrianBDouglas Год назад

      It's table 1 of the document that is in the link above it - PID Control System Analysis and Design - IEEE Article: bit.ly/2KoWMWV. I think the "table 1" bullet should have been indented or made more clear.

  • @nicholasn.2883
    @nicholasn.2883 4 года назад

    This was a great little course. Thanks

  • @sivebaartman9001
    @sivebaartman9001 6 лет назад +4

    Hi Great Video ! I was wondering what do you think about using nature-inspired heuristic methods, such as Genetic Algorithms and Particle Swarm Optimization algorithms to tune the PID values ?

  • @abdullahannaijjar9306
    @abdullahannaijjar9306 4 года назад

    Please what’s the difference between PID and QFT controllers? They all control the same Electro-hydrostatic Actuator position!
    Please any answer!

  • @lincolnwallace17
    @lincolnwallace17 3 года назад +1

    This videos are amazing! Just a question: An artificial neural network could be considered as an automatic tuning?

  • @emrevatansever15
    @emrevatansever15 2 года назад

    I believe ziegler-nichols 1st method about s-shaped output response and probably usable for real life systems.

  • @na-py3zb
    @na-py3zb 4 года назад

    Hi Brian. Thanks a lot for sharing your knowledge with us. I am working on a microgrid simulink model where voltage error is at the input of the PI controller. What result should I expect at the output of the controller? Do I need to get the value of the reference voltage by tuning the controller? Also can Ziegler Nichols method be used for all the cases?

  • @nasrinedamouche8028
    @nasrinedamouche8028 3 года назад

    Just amazing! Thanks a lot

  • @mattiasecchi815
    @mattiasecchi815 6 лет назад +2

    Hi! Thanks for the explanation! May I ask you which kind of software did you use for the "blackboard" presentation?

    • @BrianBDouglas
      @BrianBDouglas 6 лет назад +15

      I draw everything in Photoshop with a black background using a Wacom tablet. I record the screen while I do the drawing using Capto (but any screen capture program will work). Then I take that video into Final Cut Pro to speed it up and cut out any mistakes I made. Finally, I record the voiceover in Final Cut and publish the video.

    • @BrianBDouglas
      @BrianBDouglas 6 лет назад +3

      Thanks Adam! I didn't know what software Khan Academy used. I obviously copied his style though because I think it's a fun and effective way to convey information. I'll let people know of Smooth Draw next time I get this question.

  • @bopapa_1979
    @bopapa_1979 2 года назад

    What is Brian using to create these visualizations?

  • @faziljalali8948
    @faziljalali8948 4 года назад

    thanks for your nice explanation.

  • @fifaham
    @fifaham 2 года назад

    I believe that PID gains tuning should not be problem since the embedded firmware of the drone maybe able to implement such kinds of algorithms to performs Automatic-Self-Tuning. Someone must have done it somewhere around the World...!

  • @andrewmorgan5541
    @andrewmorgan5541 5 лет назад

    I'm facing a challenge of refresh rate for the feedback sensor. The temp reading refreshes every 11 seconds, the PID goes from one extreme to the other in that time period and the temperature swing is large. How can I learn how to correct this?

    • @odissey2
      @odissey2 4 года назад

      Most likely the gain in the feedback is too high. Reduce P and I terms

    • @andrewmorgan5541
      @andrewmorgan5541 4 года назад

      @@odissey2 I will try this weekend thank you sir

  • @lanakayhan1275
    @lanakayhan1275 6 лет назад

    Brian that amazing thank you so much

  • @nfkt101
    @nfkt101 4 года назад

    Can we implement machine learning for better tuning and optimizing?

  • @michlgilbertclements6178
    @michlgilbertclements6178 2 года назад

    Some auto-tuning like KMC does not work.

  • @siy0808
    @siy0808 6 лет назад

    Great!

  • @isbestlizard
    @isbestlizard 4 года назад

    ahh this was amazing thanks for explaining :D

  • @bestcakesdesign
    @bestcakesdesign 4 года назад

    Is matlab free?

    • @Amine-gz7gq
      @Amine-gz7gq 7 месяцев назад

      nope, but you can download a pirate copy

  • @sruthichandrika4120
    @sruthichandrika4120 3 года назад

    Oh great

  • @fares3651
    @fares3651 5 лет назад +7

    gg lost me this one..

  • @luismesquita8202
    @luismesquita8202 4 года назад +1

    Video on PID control:
    ruclips.net/video/71o4lFRBZnk/видео.html

  • @zaspanyflegmatyk2446
    @zaspanyflegmatyk2446 4 года назад

    3:20 , lol i felt that xD

  • @uzkanda
    @uzkanda 4 года назад +1

    Lack of actual examples and edge cases, leaves manual tuning part with very little worth.

  • @zaidhamid1622
    @zaidhamid1622 5 лет назад +2

    Great explanation. Please can you a little bit go slower ? Thx again

  • @midnight_commander
    @midnight_commander 5 лет назад +1

    *Come into a technical video*
    *complain about the technicality*
    Okay then.

  • @lailomahmughal9695
    @lailomahmughal9695 6 лет назад

    Thank you Brian . I really appreciate your work it is awesome . can you please share your email id?

  • @fredlambe7096
    @fredlambe7096 5 лет назад

    way to technical ....