Multi-start nearest-neighbour + 2-opt TSP heuristic for TRACK ordering of well connections.
More...
#include <TrackOrderingTSP.hpp>
Multi-start nearest-neighbour + 2-opt TSP heuristic for TRACK ordering of well connections.
◆ numberOfStartPoints()
| TrackOrderingTSP & Opm::TrackOrderingTSP::numberOfStartPoints |
( |
std::size_t |
numStarts | ) |
|
Set number of start points from which to explore path creation.
- Parameters
-
| [in] | numStarts | Number of start points. Note that regardless of numStarts, the algorithm will always use at least one start point–the connection that's closest to the well head defined in the constructor. Furthermore, for any particular ordering problem, the algorithm will never use more start points than the number of connections passed to order(). |
- Returns
-
◆ order() [1/2]
| std::vector< std::size_t > Opm::TrackOrderingTSP::order |
( |
std::span< const Connection > |
connections | ) |
const |
Return the permutation that puts connections in TRACK order, using default-constructed TrackCostWeights.
- Parameters
-
| [in] | connections | The connections to order. |
- Returns
- Index permutation
p such that connections[p[0]] is the heel and subsequent entries follow the optimised path.
◆ order() [2/2]
| std::vector< std::size_t > Opm::TrackOrderingTSP::order |
( |
std::span< const Connection > |
connections, |
|
|
const TrackCostWeights & |
weights |
|
) |
| const |
Return the permutation that puts connections in TRACK order, using the supplied cost weights.
- Parameters
-
| [in] | connections | The connections to order. |
| [in] | weights | Cost weights for the edge-cost function. |
- Returns
- Index permutation
p such that connections[p[0]] is the heel and subsequent entries follow the optimised path.
The documentation for this class was generated from the following files:
- opm/input/eclipse/Schedule/Well/TrackOrderingTSP.hpp
- opm/input/eclipse/Schedule/Well/TrackOrderingTSP.cpp