Dec-01-2020, 01:02 PM
Hello, I need to carry on a land-use spatial optimization where in each cell I need to assign only one type of land use. I optimize with respect two objective functions that i need to maximize, and the number of different land uses is 7. The number of cells is 1300.
My approach is to use binary variables in each cell which indicate if land use "j" is in cell "i" or not, subject to the constraint that only one type of land use in each cell can exist.
This leads to a total amount of binary variables between 4000 and 5000. This is too high for a traditional optimization solver, so i use heuristic (genetic algorithms, etc.) to approximate a set of solutions (frontier of possibilities). I have not tried it yet, so: Someone knows if this is too much computational effort for a standard computer? If so, any recommendation?
Thanks
My approach is to use binary variables in each cell which indicate if land use "j" is in cell "i" or not, subject to the constraint that only one type of land use in each cell can exist.
This leads to a total amount of binary variables between 4000 and 5000. This is too high for a traditional optimization solver, so i use heuristic (genetic algorithms, etc.) to approximate a set of solutions (frontier of possibilities). I have not tried it yet, so: Someone knows if this is too much computational effort for a standard computer? If so, any recommendation?
Thanks