Hi Sir,

I want to convert following MATLAB code into SCL can you please help me out..

function [VlvPos, FlPos, FlMove,SwPLL,SwPL,SwPH,SwPHH] = fcn(InCtrl, SPVlv, Deadband, Hyst, MinPos, kvMaxFl, kvFLsil, kvMaxVlv, kvVlVsil, SpMax, SpMin, ULInCtrl, LLInCtrl, FlNoMove, ST_Off, FlPosN_1, VlvPosN_1, t_step)

persistent MoveToMinPos RO_Flap RC_FLAP MoveToZero FlMoveUp FlMoveDown LLFl 

norenendur

Dear anahita,

 

Let’s take it step by step.

Step 1 is totally blank. If you can put it in an equation, maybe we can get somewhere. Right now, I can only agree with you unconditionally.

Step 2, „let it be!”, or W0:=0;

So, step 4 seems to be the first real issue.

To solve for dxk and lambda in that system, you need the matrices expanded and all coeficients factored out.

I haven’t done any dynamic optimization progamming so far, but from what I see, Q is already known (and probably full of scalars) and A can be computed from relationships form scalars (as in „aij=3*0 + k*i”, sort of). I believe Awk probably will be refreshed at the start of each cycle. So I would prepare the matrices in Matlab before SCL:

3.1: get the actual coeficients for Awk, probably from linear relationships of Wk.

3.2: get a big simbolic form for the entire matrix with Q and Awk. Like one full of symbolic factors, that you could use in an algebraic formula.

3.3: factor out the matrix multiplication: if dxk is a vector, get it on one side of the equation, with everything else, symbolic or scalar on the other side. Check out Matlab’s symbolic toolbox for this, they are good enough at it (though there are other packages that might be better...). You should get a (probably big) algebraic equation for each dxk and lambda. As long as it’s a single equation for each parameter, you can readly but it in SCL. It should be something like „dxki=3*i+wki+5*....”

Steps 6 and 9 should be just a series of fors. If you’re into this whole thing, I ain’t going into details. If you need them, you might be in the wrong shoes.

Step 12 might be like „for every i that doesn’t belong to Wk, the Jacobian matrix (was that the name of it?...) does something”. But those squished „less than” signes totally elude me. I’ll have to bail out on this one.

14 to 20 are just for loops, as far as I can see.

Don’t know if this is useless to you. I’m using Matlab to compute and test PIDs, feedforwards, and some MIMOs and symbolics rarely, but they have so many built in toolboxes that I’ve rarely had to hard code matrix multiplications:)!

 

Wish you luck and let me know how it goes, or if you want another second oppinion.

------------------------------------------------------------------------------------------
Split from Convert a MATLAB code.

Hello 

I want to convert following MATLAB code into SCL.. Can anyone please help me out..please..

function [VlvPos, FlPos, FlMove,SwPLL,SwPL,SwPH,SwPHH] = fcn(InCtrl, SPVlv, Deadband, Hyst, MinPos, kvMaxFl, kvFLsil, kvMaxVlv, kvVlVsil, SpMax, SpMin, ULInCtrl, LLInCtrl, FlNoMove, ST_Off, FlPosN_1, VlvPosN_1, t_step)

persistent MoveToMinPos RO_Flap RC_FLAP MoveToZero FlMoveUp FlMoveDown LLFl