Difference between revisions of "Category:Optimization"
| Line 3: | Line 3: | ||
*[http://lpsolve.sourceforge.net/ LPSolve], which is freely available | *[http://lpsolve.sourceforge.net/ LPSolve], which is freely available | ||
| − | *[ | + | *[https://soplex.zib.de SoPlex], which is freely available for academic use |
| − | *[ | + | *[https://www.gurobi.com Gurobi<sup>®</sup>], versions 6.0 - 8.1 |
| − | *[ | + | *[https://www.ibm.com/support/pages/how-do-i-download-cplex-optimization-studio?mhsrc=ibmsearch_a&mhq=cplex IBM CPLEX<sup>®</sup>], version 12.2 |
| − | *[ | + | *[https://www.mosek.com MOSEK<sup>®</sup>], version 7.1-8.1 |
| − | *[ | + | *[https://www.fico.com/en/products/fico-xpress-solver FICO Xpress<sup>®</sup>], version 7.9 |
CPLEX, Gurobi, Xpress and MOSEK are very efficient, state-of-the-art solvers. They require a license, which can be obtained for free for academic use. | CPLEX, Gurobi, Xpress and MOSEK are very efficient, state-of-the-art solvers. They require a license, which can be obtained for free for academic use. | ||
Revision as of 10:12, 10 September 2020
About the optimization tool
For PlanWise a very flexible optimization tool is built-in for formulating and solving LP and MIP problems. It is basically a graphical user interface to the ZIMPL optimization modelling language. For solving a problem, external third-party solvers are used and directly linked to the optimization tool. Currently, the solvers available are
- LPSolve, which is freely available
- SoPlex, which is freely available for academic use
- Gurobi®, versions 6.0 - 8.1
- IBM CPLEX®, version 12.2
- MOSEK®, version 7.1-8.1
- FICO Xpress®, version 7.9
CPLEX, Gurobi, Xpress and MOSEK are very efficient, state-of-the-art solvers. They require a license, which can be obtained for free for academic use.
The optimization model is linked to a Heureka-formatted SQL Server database used for storing input data, simulation data, and result data. This enables a seamless integration with the input data required by the optimization model, as well as direct presentation of optimization results in the form of tables, graphs, and maps. This simplifies the analysis and visualization of scenarios or plans of forest development and outputs.
A number of basic models will be developed with the built-in optimization tool. A user should be able to use these as they are, or as a starting point for further development.
User's guide
Available features
- Opening-size constraints can be automatically generated, resulting in a so called EARM model developed by Goycoolea et al. (2005). There is a built-in function to compute adjacency pairs, and enumeration of harvest blocks (combinations of polygons) that meet a given maximum opening size tolerance.
- Partial set-asides (hänsynsytor) within stands are handled as separate treatment units in which no harvest activities are carried out. The development in such set-asides is in reality affected by the surroundings in the way that the risk of mortality increases in the set-aside after clear-cutting the main stand. In PlanWise, you can let the program generate one alternative for the set-aside area for each management program generated for the main stand, and let the mortality be affected by when and if clear-cutting takes place in the main stand. This is non-default, by default only one alternative is generated for the set-aside. If choosing the non-default option, you must add a link between the set-aside and the main stand in the optimization problem. There is a built-in function available from the optimization menu that does this automatically. Read more here: http://heurekaslu.org/help/index.html?naturvardsatgarder.htm, section "Lämnaande av hänsynsytor" (only in Swedish).
Ongoing development (april 2014)
Another spatial feature is currently being added by Andrew B. Martin (Dalhousie University, Canada), that makes it possible to calculate neighborhood areas for each stand. This means that for a user-specified radius originating from the centroid of each stand, the area for every other stand that is overlapped by the circle is computed. This information can then be used in optimization models that take habitat area requirements into account (see {http://www.nrcresearchpress.com/doi/abs/10.1139/x10-232#.Uz5hhfmSx8E Öhman et. al 2011]).