Why do i get "Array indices must be positive integers or logical values." error?

I keep getting this error from this function. Does anyone have an idea what is really happening?

%% SIMULATION

function [sys,x0,str,ts] = ctrl_sliding_mode(t,x,u,flag)
switch flag
    case 0
        [sys,x0,str,ts]=mdlInitializeSizes;
    case 3
        sys = mdlOutputs(t,x,u);
    case {2,4,9}
        sys=[];
    otherwise
        error(['Unhandled flag = ',num2str(flag)]);
end

function [sys,x0,str,ts]= mdlInitializeSizes
sizes = simsizes;
sizes.NumContStates = 0;
sizes.NumDiscStates = 0;
sizes.NumOutputs = 3;
sizes.NumInputs = 6;
sizes.DirFeedthrough = 1;
sizes.NumSampleTimes = 0;
sys = simsizes(sizes);
x0 = [];
str = [];
ts = [];

function sys = mdlOutputs(t,x,u)

lvp_ref   = u(1);  %REFERENCE SIGNAL 
dlvp_ref = u(2);
lvp = u(3);
beta = u(4);
zeta = u(5);
dlvp = u(6);

eta = 0.2;
c  = 10;

e  = lvp - lvp_ref;
de = dlvp - dlvp_ref;
s  = c*e;

M=2;
if M==1
     ut  = 1/(c*beta*lvp^zeta)*(-eta*sign(s) + c*dlvp_ref);

elseif M==2 %Saturated function
    delta=0.2;
    kk = 1/delta;
    if abs(s)>delta
        sats = sign(s);
    else
        sats=kk*s;
    end
    ut  = 1/(c*beta*lvp^zeta)*(-eta*sats(s) + c*dlvp_ref);
end

   sys(1) = ut;
   sys(2)= e;
   sys(3) = de;

[Nithin Banka]

Can you provide the error message with all the line numbers?

[Nithin Banka]

and what are t, x, u?

[Sandro Lecci]

Hi Endrias,

Which line gives you this error?

[Endrias]

I'm tring to simulate a controller in simulink and this is what i got from Diagonstic Viewer window.

Simulation  1 Clear
08:40 AM Elapsed: 0.653 sec
An error occurred while running the simulation and the simulation was terminated
Caused by:
Error in 'sim_sliding_mode/S-Function' while executing MATLAB S-function 'ctrl_sliding_mode', flag = 3 (output), at time 0.0.
Array indices must be positive integers or logical values.

[Sandro Lecci]

I think the problem is here:

elseif M==2 %Saturated function
  delta=0.2;
  kk = 1/delta;
  if abs(s)>delta
      sats = sign(s);
  else
      sats=kk*s;
  end
  ut  = 1/(c*beta*lvp^zeta)*(-eta*sats(s) + c*dlvp_ref);
end

is sats a variable? if yes then why you want to access a specific coordinate (s)?

[Endrias]

@Sandro Sats(s) is a function like sign(x) i.e. peicewise. And when i choose e.g. M = 1, i get the next error eventhough i see that a vector of length 3 is passing put as an output...

An error occurred while running the simulation and the simulation was terminated Caused by: Output returned by S-function 'ctrl_sliding_mode' in 'sim_sliding_mode/S-Function' during flag=3 call must be a real vector of length 3;

[Sandro Lecci]

What is supposed to do the function sats? According to your if loop when M==2, sats is a variable that stores either the sign of s or the multiplication between s and kk (variables as well).

Answer 1

[Walter Roberson]

s = c*e but c=10 and e = lvp - lvp_ref but lvp_ref is scalar and lvp = u(3) . So lvp is scalar and so e is scalar and so s is scalar. That would make sign(s) a scalar, and with kk being scalar, kk*s would be a scalar. So sats is a scalar. But you have sats(s) in the ut calculation, which is an indexing request.

[Endrias]

You are prefectly right @Walter. It works now ... thank you very much all!!!