add up a signal for every timestep

hello i've got an embedded matlab function in my model, which creates an output signal(wastage of fuel). Now i want to add up this signal for every timestep, so that i obtaine the total wastage after the simulation. Is there a way? Hope that was understandable ;)

[Azzi Abdelmalek]

You want to add the signal to what?

[st]

I mean i want to sum the wastage after every timestep so that i get the total wastage after the simulation.

Answer 1

[Azzi Abdelmalek]

Ok, you want a cumulative sum:

In your embedded function add these codes

at the beginning

function [☺utput,cum_output]=fcn(....)
persistent cum_output
if isempty(cum_output)
  cum_output=0
end

and at the end of your code

cum_output=cum_output+output

[st]

okay, am i right? B is the function for wastage of fuel. A is another function.

function [B,cum_B,A]= engine(....)
persistant cum_B
if isempty(cum_B)
 cum_B=0
end
"B=....."
cum_B=cum_B+B

hope i understood this right

[Azzi Abdelmalek]

Exact

[st]

okay, i tried and get an error: "The PERSISTENT declaration cannot be used on function outputs.

Function 'Subsystem1/Embedded MATLAB Function' (#69.60.70), line 2, column 1: "persistent" Launch diagnostic report."

here is the code:

function [A,B,cum_B]= wka(q,ow,uw,eta_get,eta_gen,n,dt,P_E)
persistent cum_B
if isempty(cum_B)
  cum_B=0;
end
%#eml
A = (ow-uw)*q*eta_gen*eta_get;
Pi=P_E(end);
B=0.5*(Pi+P)*dt;
cum_B=cum_B+B;

have you an idee what is wrong?

[Azzi Abdelmalek]

It's right, try this

function [B,cum_B,A]= engine(....)
persistent cum_B1
if isempty(cum_B1)
cum_B1=0
end
"B=....."
cum_B1=cum_B1+B
cun_B=cum_B1

%And it's persistent (with e) not persistant

[st]

okay, it seems to work, at least there is no error. But strangely the difference in the scope block in comparison to before is extrem, and the outputsignal from A also changed...

function [A,B,cum_B]= wka(q,ow,uw,eta_get,eta_gen,n,dt,P_E)
persistent cum_B1
if isempty(cum_B1)
  cum_B1=0;
end
%#eml
A = n*9.81*(ow-uw)*q*eta_gen*eta_get;
Pi=P_E(end);
B = 0.5*(Pi+P)*dt;
cum_B1=cum_B1+B;
cum_B=cum_B1;

i don't see the mistake

[Azzi Abdelmalek]

I noticed that you are not working with the same code, you've posted two different codes.

[st]

thats right, i tried a code from an other nearly identical model to see if it works with it.

[st]

its strange, it seems that the signal B hasn't changed. Signal B needs signal A to be calculates(in the last code i post, P and Pi have to be A and Ai!). Could this be a problem?

[st]

another idee: I put my signal in a "sum" block, split the output from the "sum" block and return one part of the output in a unit delay block and than reenter in the "sum" block.

Hope this is understandable.

[Azzi Abdelmalek]

st, I can't answer this question, because you did not provide enough information. What is sure is the code I added allows you to do cumulative sum.

[st]

ok, understand that, even so thank you for your help.