Optimisation Systems
Our staff have advanced degrees in the theory of Operations Research, and we have added to the theory with years of practical application in this area. Recent advances in both computing power and mathematical solvers available have made a number of “too hard” problems a possibility for solution by text-book solutions thought to be too complex for a “real world” problem.
In a recent project, we solved a multi-commodity network distribution model for NZ Post that would have been impossible even 2 or 3 years ago.
While direct optimisation is not always possible, Orbit Systems is well equipped to determine whether optimisation is appropriate, and to handle this type of project.
Our capabilities include
- Linear, quadratic and non-linear programming
- Dynamic programming, including stochastic dynamic programming
- Stochastic optimisation and decision trees
- Game theoretic models
- Network modelling and routing
Example Products & Services
- Orbit Systems uses GAMS, CPLEX and Matlab for many optimisation projects.
- Dublin: A Cournot-based oligopoly model that incorporates multiple nodes and hydro reservoirs. This model was successfully used to show the Commerce Commission the effects on competition of Contact Energy’s $500 million purchase of a power station in New Zealand.