LINDO API 6.0 was released in April 2009. It includes new features to allow users to incorporate uncertainty into their optimization models. LINDO API 6.0 also includes enhancements to the linear, integer and global solvers.
New Stochastic Programming Interface
- Modeling and optimization for models with uncertain elements via multistage stochastic linear, nonlinear, and integer stochastic programming (SP).
- Extensive set of API functions to setup and solve SP models of all types, including,
- Benders decomposition for large linear SP models.
- Deterministic equivalent method for solving nonlinear and integer SP models.
- Distribution types supported: > 20, e.g., Normal, Poisson, etc.
- User defined functions allowed via call-backs.
- Customized sampling scenarios via the statistical sampling API.
Statistical Sampling API
- Extensive API functions to sample directly from various statistical distributions.
- Variance reduction with Latin-Hyper-Cube and Anti-thetic variates sampling.
- Generation of correlated samples via Pearson, Spearman, or Kendall correlation measures.
- Pseudo random uniform generation via a choice of three different generators.
Simplex Solver Improvements
- Large linear models solve an average of 20% faster with improved primal and dual solvers.
MIP Solver Improvements
- Substantial improvements in all heuristics for finding close to optimal solutions quickly.
- Significant improvements in cuts for certain types of special model structures.
Global Solver Improvements
- Significant improvement in the handling of nonlinear models with quadratic terms, especially non-convex quadratic expressions.
|
