University of Twente Student Theses

Login

Solving The Multi-Sensor Assignment Problem Efficiently

Bijl, Leander Christiaan van der (2021) Solving The Multi-Sensor Assignment Problem Efficiently.

[img] PDF
1MB
Abstract:In this report, it is investigated whether the computational cost of solving the NP-hard Multi-Sensor Assignment Problem (MSAP) can be decreased. Currently, this prob-lem is solved by the Kruger method, which first solves the problem as an assignment problem and then applies a Branch and Bound algorithm. In this report, several other methods were created that use this same solving ap-proach but try to make it more efficient and experiment with different strategies. Most of these methods gave significantly better performance than the Kruger method. However, it was also investigated whether another approach could also be used, namely solving the MSAP as a Mixed Integer Problem (MIP). The commercial pro-gram Gurobi was used to test the performance of this approach. It turned out that Gurobi performed much better on most problems than all of the earlier tried meth-ods. Because of this it is recommended to investigate the MIP approach further and see whether it can be fine-tuned specifically for an MSAP such that it also gives better performance for small problems.
Item Type:Essay (Master)
Clients:
Thales, Hengelo
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:31 mathematics
Programme:Applied Mathematics MSc (60348)
Link to this item:https://purl.utwente.nl/essays/88765
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page