Thanks a lot dear. I found lots of literature on quadrotor pid design and control. But your way of repeating is awesome. We are waiting for more lectures on different controllers and 3D version also.
Nice tutorial , if I want to implement for the 3d trajectory what will be the resultant linearised equation of motion. Can you please share the reference paper.
Hi Vin! Thanks for the tutorial! Just one question... observing the trend of the commanded input of the inner loop (M =u2), it reaches high values (hundreds of Nm)... Would it be more appropriate to consider a saturation of the inner loop too?
I tried to simulate my code using matlab but I found different values for kpz and kdz, using the same values so it's kinda weird... maybe there's something wrong in my controller.
My dude, why did you flip the y double dot and phi double dot equations around in the 1:42 slide? It breaks the visual continuity making the problem harder to study.
Greetings from Spain, this is a fantastic video, One question, in order the design the altititude and attitude controlles of the quadrotor, Is it okey to ignore the motor dinamics? That is to say, the output of the PID is a force, then the force is converted to angular velocity of the rotor, then a command pwm is generated, and finally te motor generate a force which is the one that goes into the plant (equations of motion). It seems okey simply ignoring the plant that model the actuators , don't you think?
Hi, thank you for great videos. I have a question. Why did you use saturation for outer loop ? I am designing 2 axis gimbal controller outer(position) and inner(rate ) loop. Can I have to use, to saturation for outer loop ? What is logic of saturation limits at cascaded pid ? Can you help me about this issue ? Thank you
Good Tutorial But Remember, earth radius is a vector unit. I think an equation at z should be z_ddot=-g+(u1/m)cos(theta) cos(theta)~=1 z_ddot=-g+(u1/m) please correct me if i'm wrong Thank you
Id recommend using an LQI controller in combination with an observer. Also for the Modelling of the 3D dynamics of the drone its best to use the Lagrange approach for the equations of motion. Also don't forget to take into account the linearization you have done when implementing the controller on a real system. The Controller may work with your linearized model, but if u use it on the nonlinear plant without the addition of ur stationary point where u linearized ur system around the controller will most likely fail.
Hi Vin I reproduced it, but I faced a problem in the first integrator of z in the subsystem. This is the prompt from matlab: -------------------- An error occurred while running the simulation and the simulation was terminated Caused by: Derivative of state '1' in block 'cascade/Subsystem/z' ' at time 47.808863760417289 is not finite. The simulation will be stopped. There may be a singularity in the solution. If not, try reducing the step size (either by reducing the fixed step size or by tightening the error tolerances) ------------------- I reduced the step size, but I have still the problem. I am appreciated for your assistance in advance.
I used a saturation block after the z-controller and z-output raised up exponentially, there is no any problem in phi and y. Thanks for your consideration.
@@nicolasdziemian9942 Thank you. By checking PID controller and tuning it, Matlab recommended negative p and d parameters therefore, I found out that the problem was related to the sign of (g) and (u/m) in the subsystem.
This was good till the last segment, you clearly do not understand the exercise you were following at the fundamental level. You did not add the saturation function to "stabilise much quicker", which it doesn't, it increases from 0.1 to 0.6 seconds. You added the saturation function to affirm your use of small angle approximations in the plant model. In reality this large angle would be much quicker but would affect the z output as well as the y, your plant model does not account for this, and as such you have to limit to small angle (which should really be 88 to 90, not 85 to 90 as you have)
I have bought a mini drone. And am interested in the physics and mechanics behind its flight. This lecture is exactly what I want. Thank You!
Thanks a lot dear. I found lots of literature on quadrotor pid design and control. But your way of repeating is awesome. We are waiting for more lectures on different controllers and 3D version also.
Hi Vin, Thanks for the awesome explanation.
How did you choose kp, ki and kd values and how did you mathematically model the quad rotor?
Nice tutorial , if I want to implement for the 3d trajectory what will be the resultant linearised equation of motion. Can you please share the reference paper.
God bless you man, you are awesome!
with that scheme, response of z with pid giving to me unstable output. i don't know why..
Where can I get the outputs of 8.58min?
Hi Vin! Thanks for the tutorial!
Just one question... observing the trend of the commanded input of the inner loop (M =u2), it reaches high values (hundreds of Nm)... Would it be more appropriate to consider a saturation of the inner loop too?
sir may know on which software you have modeled the aerial drone, matlab software or any other type of software
Thanks for the informative vid!
Why are we considering PID Y instead of just taking the reference value for Phi?
Thank you for this tutorial !
Hi.what's the pdf name used in the end of these toturial
thank you for this tutorial I learned a lot
I tried to simulate my code using matlab but I found different values for kpz and kdz, using the same values so it's kinda weird... maybe there's something wrong in my controller.
Very helpful! Thank you!
My dude, why did you flip the y double dot and phi double dot equations around in the 1:42 slide? It breaks the visual continuity making the problem harder to study.
Hello sir ..can you help me out in designing Quadcopter with 6DOF in simulink....
Great efforts ..Thanks
hi, kindly provide referred research paper/book etc for your mathematical model of drone.
Love your channel. Thank you for the upload!
Could you please help me to establish UAVs mathematical modeling?
Could you please eleborate why did you put u1=g and u2=g?
Greetings from Spain, this is a fantastic video,
One question, in order the design the altititude and attitude controlles of the quadrotor, Is it okey to ignore the motor dinamics? That is to say, the output of the PID is a force, then the force is converted to angular velocity of the rotor, then a command pwm is generated, and finally te motor generate a force which is the one that goes into the plant (equations of motion). It seems okey simply ignoring the plant that model the actuators , don't you think?
Its ok for this case, as long as the input-output relation is near unity (similar)
Hi, thank you for great videos. I have a question. Why did you use saturation for outer loop ? I am designing 2 axis gimbal controller outer(position) and inner(rate ) loop. Can I have to use, to saturation for outer loop ? What is logic of saturation limits at cascaded pid ? Can you help me about this issue ? Thank you
It prevents overshooting
Thank you very much for this tutorial
thanks. You give me more clarity about equation of motion?
Very nice work
can you make a 3d model for this as well ?
great!You give more background information? I don't really understand
Hi, the model didn't give me same results when I implemented in Simulink. Because of double integrators system is highly unstable. Can u guide
Hey, check the initial conditions or use a different solver
@@VDEngineering can you share your email id as my problem remains unsorted
Awesome 😎👍
Good Tutorial
But Remember, earth radius is a vector unit.
I think an equation at z should be
z_ddot=-g+(u1/m)cos(theta)
cos(theta)~=1
z_ddot=-g+(u1/m)
please correct me if i'm wrong
Thank you
Could you do videos on introduction to Simulink?
The MATLAB channel has great videos on that. I recommend you watch those!
Sir if possible please guide how to design LQR CONTROLLER for quadrotor .... during this lockdown period it vl be great help for me🙏
It's coming in a future video
Id recommend using an LQI controller in combination with an observer. Also for the Modelling of the 3D dynamics of the drone its best to use the Lagrange approach for the equations of motion. Also don't forget to take into account the linearization you have done when implementing the controller on a real system. The Controller may work with your linearized model, but if u use it on the nonlinear plant without the addition of ur stationary point where u linearized ur system around the controller will most likely fail.
Thanks man,
Thank you！
Hi Vin
I reproduced it, but I faced a problem in the first integrator of z in the subsystem. This is the prompt from matlab:
--------------------
An error occurred while running the simulation and the simulation was terminated
Caused by:
Derivative of state '1' in block 'cascade/Subsystem/z' ' at time 47.808863760417289 is not finite. The simulation will be stopped. There may be a singularity in the solution. If not, try reducing the step size (either by reducing the fixed step size or by tightening the error tolerances)
-------------------
I reduced the step size, but I have still the problem.
I am appreciated for your assistance in advance.
I used a saturation block after the z-controller and z-output raised up exponentially, there is no any problem in phi and y.
Thanks for your consideration.
@@alirezasharifan9663 It sounds like it's something within your gain values within the PID Controller
@@nicolasdziemian9942 Thank you.
By checking PID controller and tuning it, Matlab recommended negative p and d parameters therefore, I found out that the problem was related to the sign of (g) and (u/m) in the subsystem.
You can be a math nerd but its all about the skills of flying the drone.
Trust me I'm a good pilot.
Why need skills when he can design the controller
@@minimon796 because developing your own skills is not a cheating.
@@federicosagun8243 Flying drones with hands requires no skill at all. Designing an auto controller however, do require special knowlege and skills
@@minimon796well it is much fun and very challenging when flying the drone with full authority and control. Drone pilot prefer it that way.
What about 3D controller
coming in future
@@VDEngineering future when :(
Give me report.pdf, please...
Doenst work
This was good till the last segment, you clearly do not understand the exercise you were following at the fundamental level.
You did not add the saturation function to "stabilise much quicker", which it doesn't, it increases from 0.1 to 0.6 seconds.
You added the saturation function to affirm your use of small angle approximations in the plant model. In reality this large angle would be much quicker but would affect the z output as well as the y, your plant model does not account for this, and as such you have to limit to small angle (which should really be 88 to 90, not 85 to 90 as you have)