I was using MATLAB to automatically construct and build mathematical models from blocks. After converting the model from state space to transfer function, I found that step responses are not close to each other, why?
The m-file is included.
Hi your numerator and denominator are incorrect from the output of ss2tf().
I solved it by hand from the block diagram and verified by plotting that your overall system transfer function needs to be:
Tf(s) = 160/(s^2+2*s+160)
The real question is why is ss2tf() not returning the correct num and den?
See attached...
You might find this helpful
http://www.mathworks.com/matlabcentral/answers/9445-ss2tf-and-tf2ss-function
Hi your numerator and denominator are incorrect from the output of ss2tf().
I solved it by hand from the block diagram and verified by plotting that your overall system transfer function needs to be:
Tf(s) = 160/(s^2+2*s+160)
The real question is why is ss2tf() not returning the correct num and den?
See attached...
You might find this helpful
http://www.mathworks.com/matlabcentral/answers/9445-ss2tf-and-tf2ss-function
Markus
The command tf with state space model as an input returns the correct transfer function that would appear by manually reducing the block diagrams (i.e., the one found by William). Both responses match each other now.
I think that MathWorks should point out that ss2tf command does not work if E matrix is not an identity matrix in the state space model.
Thanks all
What i think is happening in the state space form is that matlab gives the answer in mormalize form. try and and create simulink block for both the transfer function and state space and run them with step input
- Badges
- Science topic