A pure java MIP Solver perfect for hard NP Problems.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Libraries Or Mar 5, 2014
QBnBJava Or Mar 5, 2014
README.md Update README.md Mar 9, 2014

README.md

QBnB

A pure java MIP Solver perfect for medium size NP Problems.

State-of-the-art solvers like Gurobi, CPLEX and XPress are fast, robust, but expensive. The idea behind this project is to develop a complete set of optimization libraries in Java for making embeded Decision Support Systems (DSS) for commercial and academic use.

The final goal is to develop the following capabilities in the project:

Exact Algorithms

  1. A fast LP Solver with the Simplex Algorithm implemented in standard and tableau format.
  2. A pool of polynomial time interior point methods for LP.
  3. A MIP Solver that implements Branch and Bounding, Branch and Cut, and Branch and Price.
  4. Column Generation
  5. Dantzig-Wolf Decomposition
  6. Benders Decomposition
  7. Network efficient algorithms (e.g. Dijkstra)

Metaheuristics

  1. A pool of metaheuristics, and mateheuristics for combinatorial problems.

Specific frameworks

  1. A Object-Oriented framework for solving VRPTW, CVRP, and other variants of the Vehicle Routing Problem.
  2. Supply Chain problems

if you are interested in this project fork us! We encourage Operations Research students to contribute to this project.

Contact: Juan Imbett, jfimbett@gmail.com

Current Work:

The following modules are in their beta test:

  1. LP and BLP problems
  2. MIP using Deep Search First and Branch and Bound