Issue with Model Feasibility in CPOptimizer 22.1

Hi ​@Serval2501. There is no X(g) that can be 1 and satisfy the constraints XX2, DistMin_Points and DistMin_Centres. You can see this as follows:

1. Change the definition of the objective variable MaxResult into: sum( g, X(g) ).
2. Deactivate constraint Max_NC by adding "| 0” to the index domain.
3. Solve the model.

You will see that the objective value is 0, so the maximum value for sum( g, X(g) ) equals 0, which implies that none of the X(g) can be 1.

Note that constraint XX2 enforces that X(g) = X2(g) for all g, so you could replace X2 by X in your model, and delete X2 (unless XX2 contains a mistake).

The original definition of your model contains several min() terms which is quite "ugly” and makes the model nonlinear. The min( Dist_Centre(g), … ) term evaluates to Dist_Centre(g) because Dist_Points(g, g_bis) = 0 if g=g_bis, so in its current form you do not need the min() terms.