State space control - observer design using Matlab and Simulink

This video is intended to help you understand implementation a linear observer in a Matlab/Simulink environment.
I invite you also to see these nice videos:
A Conceptual Approach to Controllability and Observability: • A Conceptual Approach ...
and
Understanding Kalman Filters, Part 2: State Observers:
• State Observers | Unde...

[@gavindundee1317]

You are amazing! Thank you for this extremely high quality and easy to understand video.

[@ogunsadebenjaminadeiyin2729]

This video is very good, thanks for sharing prof. Very very very nice video:

[@johnjoseph6634]

Awsome...........loved your content. make more videos sir

[@ketz108932]

Nice, Well eplained

[@mahmadkhachfe9089]

❤

[@sagarjha6043]

The video is amazing

[@josephkessel9535]

Hi! Dear, thanks a lot for this helpfull video on linear observer. Pls, can you also have some matirials which can help to design an obsever with disturbance rejection with a state feedback control law? Thanks a lot.

[@vighneshshenoy1462]

Thanks a lot!!! can u please do a video on non-linear time variant system observer design.

[@ahmadhably80]

I have to prepare something for September

[@ahmedelnahas8933]

Hello Mr Ahmed , I'am working on a MBBR ( Micro ball balancing robot ) system and i already design the control for the system but i don't know how to choose the poles of the observer which will satisfy my system and gives the appropriate state error can u guide me ?

[@ahmadhably80]

Of course. You can email me please with your problem

[@anne-sophieg4488]

Hello Mr Ahmad, First of all, thank you for your video. It helped me a lot! But I have a question, why the system is diverging at 4:26? Thank you in advance for your answer. Have a great day!

[@ahmadhably80]

Because the system is already unstable. Please see the Eigen values of A

[@anne-sophieg4488]

@@ahmadhably80 Thank you for your quick answer! Indeed, the system is instable. So on the next part of the video, you changed the value of A? Thank you advance for your answer.

[@ahmadhably80]

No I am showing that the error between the estimated state and the real state is converging to zero

[@anne-sophieg4488]

So it is possible to create an observer even if the system is diverging ?

[@anne-sophieg4488]

Thank you!

[@theyeah132]

How did you get the state vector from the state space blocks? Mine only outputs y!

[@ahmadhably80]

The matrix C must be identity and D has to zeros of an appropriate size

[@theyeah132]

@@ahmadhably80 Ah of course! Thank you for the help, the videos are awesome by the way.

[@ahmadhably80]

Thank you. Please spread the word

[@younesshmd9852]

@Ahmaad Halby Thank you Professor for this course, can you please tell me how I can make a discreet time luenberger observer? I can send you with your permission the problem that I have by email.

[@mohammedabbad5938]

I have DC-DC converter and I need to design observer based-controller for the converter I used PID controller to control the output voltage to be be same as the reference voltage but I don't know how to design the observer

[@ahmadhably80]

Can you give your system’s equation?

[@mohammedabbad5938]

@@ahmadhably80 The state space you mean sir ?

[@chamounmichel3602]

Thank you Prof. for this video. Unfortunately, when I run the simulation i get different results than yours. In fact, if i extend the simulation time to 30 seconds, the observer will diverge COMPLETELY! what could be the problem? (i am using the state space representation bloc)

[@ahmadhably80]

Could you share your code with me?

[@chamounmichel3602]

@@ahmadhably80 yes sure can you please share your email with me? Or if you prefer not to share it publicly, please send me an email to [e-mail]

[@ahmadhably80]

[e-mail]

[@ahmadhably80]

I examined your code. Remember an observer is a software sensor. if your system is unstable, the observer will follow the system and show this instability. use a stable system instead. Use this A = [-2 3;-1 -4]; It will work :-)

[@chamounmichel3602]

@@ahmadhably80 ah i see! Thank you!

[@gorillacapers]

I followed all the steps in the StateSpace representation, but my errors are very large; to the order of 10^11 within the first 10 seconds My matrices are the same as yours. What could be going on?

[@ahmadhably80]

A sign problem?

[@gorillacapers]

@@ahmadhably80 Signs are same as you have in your video. I double checked. The Step function Step-time is 1. When I change it to 0 I see a difference too. Either case my plots do not match yours.

[@ahmadhably80]

Send me your files

[@yizhou5230]

Does anyone know why the scope has two lines? What's the difference between blue line and yellow line.

[@ahmadhably80]

On the scope we show the system's real state and the observed state from the observer

[@sagarjha6043]

Hi! I have a system. Can you help me for designing observer matrix.

[@ahmadhably80]

?,

[@ahmadhably80]

An answer on a question nothing is going wrong. You have just to change the solver to a "stiff" one. The idea is to have a small error between the real and estimated state. The system is already unstable. If it is diverging, the observer will follow it.

[@Cghxjccy]

Plz what is matrix D Zeroes ( 2 .. What’s after 2

[@ahmadhably80]

Lust respect the number of outputs and inputs

[@zemalizakaria7694]

What is the name of this observer

[@ahmadhably80]

Luenberger Observer

[@zemalizakaria7694]

@@ahmadhably80 thanks very much teacher please Would U give us un example of observer filter kalman and unknown input if U can of cours

[@zemalizakaria7694]

Teacher please can U give me the différence between thees observers

[@akibiftakher9892]

@Ahmed Hably Please check your email.

[@ahmadhably80]

Here is a part for the correction: A11 = [-2 3; -1 -4]; B11= [0; 1]; C11 = [1 0]; D11 = 0; L1 = acker(A11',C11',[-1;-1])' Aobs1 = [A11-L1*C11]; Bobs1 = [B11 L1]; Cobs1 = eye(2); Dobs1 = zeros(2,2); For the Simulink I cannot open it since I have an older version

[@akibiftakher9892]

@@ahmadhably80 Thank you