write Simulink block diagram to source code

How to write this in matlab source code instead of using Simulink?

[Walter Roberson]

Are those transfer functions? If so, see tf()

[Unkown]

Yes. okay

[Walter Roberson]

https://www.mathworks.com/help/control/ref/tf.html

and the other models listed at https://www.mathworks.com/help/control/fixed-coefficient-models.html?s_tid=CRUX_lftnav

Answer 1

[Pravarthana P]

Hi Olivia, it can be understood from the figure and your query that you are trying to write MATLAB code equivalent to the closed loop mentioned above.

To begin with try reducing the first block to have a common denominator. Then it will be easy to code the transfer function in MATLAB using the tf() function using the following link Transfer function.

For example, I am assuming first block as G1 and second block as G2:

G1 can be re-written as:

This can be written using tf()as:

G1 = tf([ ],[ 0 0]).

G2 can also the reduced in the same manner. Assign

G=G1*G2

Next step will be to convert the closed loop to code, you can find information related to closed loop in the following link: Using FEEDBACK to Close Feedback Loops - MATLAB & Simulink Example (mathworks.com)

Try using the feedback() function, and the appropriate G and K values (in your example K=1).

To help with the process try the following documentations:

• Using FEEDBACK to Close Feedback Loops - MATLAB & Simulink Example (mathworks.com)
• Feedback connection of multiple models - MATLAB feedback (mathworks.com)
• Basic Models - MATLAB & Simulink (mathworks.com)

[Walter Roberson]

Be careful though: Simulink will not accept the first of those transfer functions

syms K_c2 K_p2 s tau_D tau_I tau_p2 theta_p2
G1 = collect(K_c2 * (1 + 1/(tau_I * s) + tau_D * s), s)

G1 = 

Notice that the numerator has s powers up to but the denominator has at most s . That situation is not permitted for transfer functions within Simulink (the documentation says something about it being non-casual when it happens.)

G2 = K_p2 * exp(-theta_p2*s)/(tau_p2 * s + 1)

G2 = 

H = G1 * G2

H = 

[N, D] = numden(H)

N = 

D = 

num = N

num = 

den = collect(D, s)

den = 

Okay, so the composite function H has up to in the numerator and also in the denominiator, and that is acceptable.

The in Laplace Transform theory matches to a delay of units.