Originally posted by: CA_19

Hello,

As I am new in IBM Cplex optimisation, I have a problem with implementing data from excel and using it to solve a problem.

I have a very large data set of nodes (x and y coordinates) as well as links in excel and I am trying to find the shortest path (and lowest cost) between a set of x nodes and another of y nodes chosen from the overall data set and considering the possible links. I also have to account for the cost of each node in the calculation.

Could you please help me with how to input the data and select the corresponding nodes from the overall matrix I am working with?

Regards,

Markus

Hi,

some links:

https://www.ibm.com/developerworks/community/forums/html/topic?id=f58daf79-f752-48e4-8f25-54b2f6f71b40

for TSP

https://www.ibm.com/support/knowledgecenter/SSSA5P_12.9.0/ilog.odms.ide.help/OPL_Studio/opllang_quickref/topics/tlr_opl_SheetRead.html

for Excel connection in OPL

regards

https://www.linkedin.com/pulse/making-decision-optimization-simple-alex-fleischer/

Originally posted by: CA_19

Thank you very much for the pointers.

However, the main difficulty is how in Opl to select the nodes that I want to set as origin and destination from the big nodes data set and link them with the objective function to minimize the (flow x cost) expression.

I have started with the tuple format and I am using a Boolean variable:

tuple node {
    string name;
    float Xcoord;
    float Ycoord;
    float InOutFlow;
}

tuple link {
    string origin;
    string destination;
    float delay;
    float length;
}

{link} LINKS = ...;
{node} NODES = ...;
 

dvar boolean Flow[LINKS];

dexpr float TotalCost = sum(b in LINKS) (10*b.length) * Flow[b] + sum(b in LINKS)sum(b in LINKS) (2*b.delay)*Flow[b];

minimize TotalCost;

but I am really confused about how to consider the specific nodes I need to analyse in this case...

Thanks

what about that example

https://www.ibm.com/developerworks/community/forums/html/topic?id=01efe4be-661e-49e2-b213-c44d16a35c21

?