Output argument 'S1' is not assigned on some execution paths.

Hi.

I'm trying to implement Space Vector PWM. I created a function to calculate my time and as well as the switching time for the inverter.

I however get "Output argument 'S1' is not assigned on some execution paths." as the error. I have looked through and try out diffrent approach, nothing seems to work.

Please i need help to resolve this persistent error.

I have also attached a copy of the file. (downgarded to 2013version)

This is the complete Matlab function.

% I have 3 input into the function block%
% The Magnitude, the Angle (radian) and the Sequence block in that order%
% The output goes to my inverter IGBT's swith input%
function [S1,S3,S5,S4,S6,S2] = fcn(u)


b   = sqrt(3);
Vdc = 400;
Fs  = 5000;
Ts  = 1 / Fs;
m   = pi/3;

Alpha = u(2);
A =  b * Ts * u(1) / (Vdc);

 if (Alpha > 0) && (Alpha <= m)
    
    K =1; %for sector 1%
    
    Ta = A * sin((K * m) - Alpha);
    Tb = A * sin(Alpha);
    To = Ts - Ta - Tb;
    
    
    a = To/4;
    b = Ta/2;
    c = Tb/2;
    d = To/2;

    T = [a b c d c b a];
    
    T_T = cumsum(T);
    
    SA = [0 1 1 1 1 1 0];
    
    SB = [0 0 1 1 1 0 0];
    
    SC = [0 0 0 1 0 0 0];
    
    for e = 1:7
        
        if(u(3) < T_T)
            break
        end
    end 
 
S1 = SA(e);
S3 = SB(e);
S5 = SC(e);

S4 = 1-SA(e);
S6 = 1-SB(e);
S2 = 1-SC(e);

 end

Answer 1

[Image Analyst]

Have the first lines right after the function line be assignments for the S's.

S1 = nan; % or zero or -inf or whatever you want.
S3 = nan;
S5 = nan;
S4 = nan;
S6 = nan;
S2 = nan;

That way, not matter what happens (like you don't assign them afterwards for whatever reason, which is what happened), they will at least have a value and that particular error won't happen.

[Abdulwasiu Yusuf]

Thanks.

I tried this approach, the error was gone, but the output was'nt as expected. it seems this line in the code

S1 = SA(e);

S3 = SB(e);

S5 = SC(e);

S4 = 1-SA(e);

S6 = 1-SB(e);

S2 = 1-SC(e);

affects the desired final output of [S1:S6] when you give it nan or -inf initialization.

[Abdulwasiu Yusuf]

Hi , Thanks so much , it finally worked

I declared it as

S1 = nan;

S3 = nan;

S5 = nan;

S4 = nan;

S6 = nan;

S2 = nan;

the other error I didn't catch was my for loop.

for e = 1:7

if(u(3) < T_T)

break

end

end

it was missing the the index checker (e). in the "if statement

I thus resolved it by changing the code to

for e = 1:7

if(u(3) < T_T(e))

break

end

end

'"

Everything now works great. !!!!

Thanks

Answer 2

[Walter Roberson]

When

if (Alpha > 0) && (Alpha <= m)

is false, then you do not assign any value to S1.

There might be other constraints in the model that force u(2) to always be in the range eps(realmin) to pi/3 so it might be a problem more in theory than in practice, but the optimizer doesn't know anything about that.

[Abdulwasiu Yusuf]

Yes, u(2) is a repeated sequence tied in the range of [0 pi/3 2pi/3 0 -2pi/3 -pi/3 0] in that order, and it corresponds to a sector in the Space Vector plane.

i.e,

0 to pi/3 Sector 1

pi/3 to 2pi/3 Sector 2

2pi/3 to 0 Sector 3

0 to -2pi/3 Sector 4

-2pi/3 to -pi/3 Sector 5

-pi/3 to 0 Sector 6

And the function block will check u(2) state and generate the time interval for each sector to occur.

Answer 3

[Abdulwasiu Yusuf]

Resolved and now working!!!

% I have 3 input into the function block%

% The Magnitude, the Angle (radian) and the Sequence block in that order%

% The output goes to my inverter IGBT's swith input%

function [S1,S3,S5,S4,S6,S2] = fcn(u)

S1 = nan;

S3 = nan;

S5 = nan;

S4 = nan;

S6 = nan;

S2 = nan;

b = sqrt(3);

Vdc = 400;

Fs = 5000;

Ts = 1 / Fs;

m = pi/3;

Alpha = u(2);

A = b * Ts * u(1) / (Vdc);

if (Alpha >= 0) && (Alpha < m)

K =1; %for sector 1%

Ta = A * sin((K * m) - Alpha);

Tb = A * sin(Alpha);

To = Ts - Ta - Tb;

a = To/4;

b = Ta/2;

c = Tb/2;

d = To/2;

T = [a b c d c b a];

T_T = cumsum(T);

SA = [0 1 1 1 1 1 0];

SB = [0 0 1 1 1 0 0];

SC = [0 0 0 1 0 0 0];

for e = 1:7

if(u(3) < T_T (e))

break

end

end

S1 = SA(e);

S3 = SB(e);

S5 = SC(e);

S4 = 1-SA(e);

S6 = 1-SB(e);

S2 = 1-SC(e);

end

end