Scribd
Upload
29 views6 pages

Automatic Control Lab 3

wqdef

Uploaded by

moha amro
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
29 views6 pages

Automatic Control Lab 3

wqdef

Uploaded by

moha amro
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Mechatronics Engineering Department

Experiment Name:

System
identification

Name: Mohamed Alamrouni


No: 3787

Mohamed alamrouni 3787


Background:

System identification is a method used to gather information


about a system by analyzing measured input-output data. The
main outcome of this process is an identification model, which
can be static or dynamic, deterministic or stochastic, and can
take on linear or nonlinear forms. These models are valuable
for various tasks, including simulation, controller design, and
system analysis.

In addition to creating models, system identification can


extract other crucial information about a system. For example,
it can identify the controllability matrix and the observability
matrix, as well as calculate gains for observers and Kalman
filters. It can also evaluate existing feedback controller gains,
analyze disturbance environments, and detect failures in
actuators and sensors.

Moreover, this methodology allows for the direct synthesis of


feedback or feedforward controller gains using input-output
data, bypassing the need for an intermediate model. This
versatility makes system identification a powerful tool in many
engineering applications, enhancing system performance and
reliability.

In this experiment, I will create state space models for a wind


turbine, ensuring that the results are sufficiently accurate. I
will also investigate how the order of the system impacts the
accuracy of the simulations.
Experiment Design and Results :

The results of the system identification experiment reveal a comprehensive continuous-


time state-space model with meaningful parameter estimates. The identified A, B, C,
and D matrices capture the complex dynamics and interactions within the system,
providing a detailed representation of the underlying processes. The relatively high fit
of 70.29% to the estimation data suggests the model is a reasonably accurate
representation of the system behavior.

However, the FPE of 8.429 and MSE of 17.73 indicate there is still room for improvement
in the model's predictive performance. Potential next steps could involve further
refinement of the model structure, exploration of alternative estimation methods, or
incorporation of additional data to enhance the model's reliability and generalization
capabilities. Overall, the results demonstrate the value of the system identification
approach in developing a data-driven model for understanding and predicting the
system's dynamics
Results:

ss2 =
Continuous-time identified state-space model:
dx/dt = A x(t) + B u(t) + K e(t)
y(t) = C x(t) + D u(t) + e(t)

A=
x1 x2 x3
x1 -0.008979 0.06139 0.1464
x2 -0.1064 0.3071 2.497
x3 0.04977 -1.77 -0.3072

B=
u1
x1 -0.0005244
x2 -0.006779
x3 0.003873

C=
x1 x2 x3
y1 179.1 10.45 11.7

D=
u1
y1 9.479

K=
y1
x1 0.004783
x2 -0.0003436
x3 0.0002936

Name: ss2
Parameterization:
FREE form (all coefficients in A, B, C free).
Feedthrough: yes
Disturbance component: estimate
Number of free coefficients: 19
Use "idssdata", "getpvec", "getcov" for parameters and their uncertainties.

Status:
Estimated using SSEST on time domain data "mydata".
Fit to estimation data: 70.29% (simulation focus)
FPE: 8.429, MSE: 17.73
Conclusion:

The system identification results have produced a


comprehensive continuous-time state-space model that
captures the complex dynamics of the studied system.
The identified parameter matrices reveal the underlying
relationships between the states, inputs, and outputs,
offering valuable insights. The model's 70.29% fit to the
estimation data suggests reasonably good performance
in simulating the observed behavior. However, the higher
FPE and MSE values indicate room for improvement in
the model's predictive accuracy. Potential enhancements,
such as exploring alternative model structures or
incorporating additional data, could further refine the
model's reliability and expand its applications in various
domains. The developed model provides a solid
foundation for understanding and simulating the
system's dynamics.
Reference:

https://lucris.lub.lu.se/ws/files/48192977/TFRT_7011.pdf

https://www.youtube.com/watch?v=u7hJ1aF-JrU

I used AI to get suggestions for this report specific the


background. It helped enhance my grammar and clarity,
making my ideas more effective.

You might also like