why is there 2 different types of output responses in the 2 simulink models of the closed loop transfer function,for the same step input given below?

4 views (last 30 days)
To be more elaborate, my question is that, why am i getting 2 different types of responses for the same closed loop transfer function when i am considering 2 cases? In both the cases i am giving step input. Below i have attached 2 simulink files i.e. "response.mdl" & "munnu.mdl" along with their respective outputs. Also in both the cases, the load torque is assumed to be zero. In the first case i have taken the total combined transfer function, & in the 2nd case i have divided the total transfer function into 2 transfer functions which are associated with their constants. Considered parameters for the transfer function are: Kt=motor torque constant=0.8 NM/A Kb=back emf constant=0.8 VS/Rad B=coefficient of friction=0.0167 J=moment of inertia=0.0167 KGM^2/S^2 Ra=armature resistance=0.6 ohm La=armature self inductance=0.012 Henry

Answers (2)

Arkadiy Turevskiy
Arkadiy Turevskiy on 9 Dec 2013
Edited: Arkadiy Turevskiy on 9 Dec 2013
The response are different for a couple of reasons:
1.
The main reason is the initial conditions are different. In response.mdl the system is initialized at zero, but in the mannu you initialize both subsystems at 0.5.
2.
Closed loop transfer functions are different. In the munnu file close loop transfer function is:
3992
-----------------
s^2 + 51 s + 3243
which is different from closed loop transfer function in response.mdl
4000
-----------------
s^2 + 50 s + 3250
If you change initial conditions to 0 in mannu.mdl, then responses become quite close. I also got rid of sampling time and reduced max simulation step to 0.001.

Balaji Rajagopalan
Balaji Rajagopalan on 25 Sep 2014
Hi,
I am facing a Similar Issue.
The Step response of a First order System, is being analyzed in two different scenarios. 1. With First order transfer function of the System, & 2. With time-domain equivalent of the same. And here also, I am getting Different Output responses, for the same step input as shown in screen capture and in the model attached.
Explanation:
With System transfer function as: 1/(s+β); (note: here β is set to 0.01) And with input step as: α/s; (note: here α is the magnitude of the Step, and is set to 20 from time t=0 to t=500, and then to 10 from time t=500 to t=100, where total simulation time is 1000.)
We get the output as: α/(s*(s+β));
Which when converted to time domain results in a following expression: y(t)= (α/β)*(1-e^(-βt)) Where “α” denotes step magnitude, and the above expression describes the change in output for a step input.
Problem:
In my Simulink Model, I designed a Sub-System as per the above time-domain equation with all the available matematical blocks; please verify my attached model (Sub-System in Pink). As per my understanding it should be the same as the step response of the 1st order transfer function (in green), right????
Now my problem is when both the Systems are excited by the same step Signal of same magnitude I get different responses as shown in the Graph attached:
The response graph in Yellow indicates the response of the transfer function, and the response in green indicates response of the Sub-system designed.
I am unable to understand this disparity. I would be grateful to you people, if you can clarify / help me on this .
Thanks in advance!!
With best regards
Balaji

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!