Originally posted by: CDN3_Daniel_Cramer

Dear Community,

I am working on my Master-Thesis and I got a problem to implement a model of the capacitated lot sizing problem. The backgroung: I try to optimize the purchasing behaviour of a company.

My problem is that I got the following message: "Operator not available for int => int"
The problemmarker says, it has something to do with the expression in the first constraint after the double dots ("b => 1")

The identifiers are all in German - sorry for that...

Code of the .mod:

float h = ...; 

int p = ...; 

int s = ...; 

int n = ...; 

int M = ...;   tuple Material 
{  string Nummer; 

int          Anforderungsmenge;
};   
{Material
}    Bedarf = ...;   
{string
} Nummer            = 
{a | <a,b> in Bedarf
}; 
{

int
}    Anforderungsmenge = 
{b | <a,b> in Bedarf
};   

int Bedarfgesamt[Nummer][Anforderungsmenge];   dvar int+     Bestand[Nummer][Anforderungsmenge]; dvar int+     Bestellmenge[Nummer][Anforderungsmenge]; dvar 

boolean  Bestellung[Nummer][Anforderungsmenge];   minimize sum(a in Nummer, b in Anforderungsmenge) (s * Bestellung[a][b] + h * Bestand[a][b] + p * Bestellmenge[a][b]);   subject to
{ forall(a in Nummer, b in Anforderungsmenge : b => 1) Bestellmenge[a][b] == Bedarfgesamt[a][b] - Bestand[a][b-1] + Bestand[a][b];   forall(a in Nummer, b in Anforderungsmenge) Bestellmenge[a][b] - M*Bestellung[a][b] <= 0;   forall (a in Nummer) Bestand[a][0] == Bedarfgesamt[a][0]; 
};

Code of the .dat:

n = 23; M = 6000; s = 100; h = 0.2; p = 10; Bedarf = 
{ <
"123",0>, <
",20>, <",40>, <
",0>, <",20>, <
",0>, <",20>, <
",20>, <",120>, <
",0>, <",20>, <
",20>,<",1120>, <
",2>, <",20>, <
",230>, <",20>, <
",20>, <",20>, <
",20>, <",20>, <
",20>,<",0>, <
",20>, <",40>, <
",0>, <",20>, <
",0>, <",20>, <
",20>, <",120>, <
",0>, <",2320>,<
",20>, <",1120>, <
",2>, <",20>, <
",230>, <",20>, <
",320>, <",520>, <
",420>, <",20>, <
",120>}; 
};

The model express for the company, in which period an order should be done and also the demand of the product (there are 2 products (123 and 456)).
I also attached the mathematical model (the variables in the model do not match with those in the OPL-model, but the structure of the variables are nearly the same).

Maybe someone has a tip for me? Or should I explaine more? (Sry for my bad English)

Thanks for your help!

Greetings from Germany
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: CDN3_Daniel_Cramer

I made a problem when I copied the .dat-Data:

The post is not correct, when I posted it.

For information look in the attached file.

That should be correct now...
#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi,

can you try to replace

b => 1

by

b >= 1

?

regards
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: CDN3_Daniel_Cramer

Life could be so easy...
That works - perferct - but I got another problem:
In the same constraint, the problemmarker says, that CPLEX is unable to extract this expression.

Last but not least - I am not sure, whether my variable-declaration (and the tuple) are correct.

If I made -- instead of your tip -- the following:

forall(a in Nummer, b in Anforderungsmenge : b == 1) 
        Bestellmenge[a][b] == Bedarfgesamt[a][b] - Bestand[a][b-1] + Bestand[a][b];

then there is no problem, but CPLEX didn't want to solve it (my decision variables are all zero)

Greetings
#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi

forall(a in Nummer, b in Anforderungsmenge : b >= 1) 
        Bestellmenge[a][b] == Bedarfgesamt[a][b] 
        //- Bestand[a][b-1]
         + Bestand[a][b];

works fine too.
The problem in your code is that Bestand[a] b-1 does not exist sometimes.

Maybe what you need is

forall(a in Nummer, b in Anforderungsmenge : b >= 1) 
        Bestellmenge[a][b] == Bedarfgesamt[a][b] 
        - Bestand[a][prev(Anforderungsmenge,b)]
         + Bestand[a][b];

Regards
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: CDN3_Daniel_Cramer

Thanks again for your comment - I tried your code, but I am not able to define the prev-command like in the OPL-documentary...
This will be the last time that I am asking you ;)

Thanks again, Mr Fleischer
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: CDN3_Daniel_Cramer

I used an OPL-example to learn, how to solve my problem!
#DecisionOptimization
#OPLusingCPLEXOptimizer