Originally posted by: c-c-blum

Hello everyone!

I would need to know how to solve the TSP with CP Optimizer using C++. Can someone point me to an efficient model please?

Best regards!

Hi,

could https://www.ibm.com/developerworks/community/forums/html/topic?id=5efb1db1-c642-4877-af60-b330f959a534&ps=25

help ?

This shows you some ideas in OPL but you could do the same with C++

regards

Originally posted by: c-c-blum

Hi! Thanks for your answer. Yes, this could definitely help.

Is there a way to translate OPL models into the corresponding C++ code? (I only know how to program CP optimizer in C++). Otherwise, I will have to learn OPL.

Thanks in advance!

Hi,

all OPL objects hide concert objects so you could either translate to C++ or call OPL from C++.

You have some examples in

CPLEX_Studio1271\opl\examples\opl_interfaces\cpp

regards

Originally posted by: JorisK

The common way to implement this problem is as follows:

1. Define an interval variable for every city. Designate one of the cities the 'starting city'. Introduce a copy of this city as the 'ending city'. So a tsp solution would be: [start_city, 3,2,4,5, end_city]. In this example, the distance from city 5 to end_city would be equal to the distance from 5 to the start_city.

2. Create a no-overlap sequence of the interval variables. This constraint also takes a matrix containing the distances between every city. Use 'first(start_city,seq)' and 'last(start_city,seq)' constraints to enforce that these 2 interval variables are resp the first and last in your sequence.

3. Specify the objective function: minimize endOf(end_city)