Browsing by Author "Strandberg, Jani"

Now showing 1 - 2 of 2

Euroopan taloudellisesta yhtenäisyydestä
(2016-04-16) Strandberg, Jani
Perustieteiden korkeakoulu | Bachelor's thesis
Semi-automated scenario analysis of optimisation models
(2019-10-22) Strandberg, Jani
Perustieteiden korkeakoulu | Master's thesis
Mathematical optimisation models have been succesfully used for solving problems across multiple industries. Often, the purpose of these models is to provide decision support by guiding the development of effective plans and decisions. In the process of providing decision support, obtaining the numerical solution of an optimisation model is usually not limiting, and increasingly large and complex models are solvable as a result of increased computation resources and improved solution algorithms. However, analysing and understanding the model results can be difficult for large and complex problems. Analysing model results often involves processing one or several model scenarios with varying parameters, so that related conclusions become available to the user. This part of the mathematical programming process is usually done by the analyst on a case-by-case basis, but this could be aided through computer assisted tools known as Intelligent Mathematical Programming Systems (IMPS).\\ In this Thesis, we develop a method for analysing and comparing the results of optimisation scenarios. This method forms a basis for a new form of IMPS that can be used to analyse both individual model scenarios and differences between two scenarios. The method is based on the idea of preserving the structure of the mathematical model along with the optimisation results, by representing the results as a graph. While the use of the method is considered in the context of supply chains and linear programming, the approach is fairly general and could be applied in other types of optimisation problems as well. The implementation of the IMPS is done with open-source technologies and can be coupled with any modelling environment and solver.\\ The usability of the developed method for scenario analysis is evaluated through a case study related to an existing supply chain optimisation model at a large pulp and paper company. We identify some scenario related questions where the developed method has advantages over the traditional approaches where the model structure is not explicitly preserved. The case study illustrates that the developed method has many potential use cases, especially in conjunction with other methods. Furthermore, multiple development needs and avenues for further study are identified.