The combustion of fossil fuels and the resulting emission of greenhouse gases (e. g. CO2) in the transport sector account for a major portion of global resource depletion and climate change. That is why, governments around the world are enacting regulations to cut CO2 emissions in road transport. These new regulations pose challenges to automotive manufactures as well as infrastructure operators that need to develop new cars with alternative powertrains and provision alternative fuels and energy sources, respectively.
The demand for different alternative fuels must be met by a corresponding supply. However, the infrastructure required for this is not in place. The development of this infrastructure is often characterized by large investments, while future demand is uncertain. Installing alternative refueling systems on already existing conventional refueling stations can mean significantly lower investments. However, this also requires the dismantling of conventional refueling systems, as space at refueling stations is limited. The operators of the existing infrastructure must therefore decide when, where and in what capacity which system is to be installed. In order to support infrastructure operators in their decisions in the transformation process, a mathemathical optimization model was developed and tested on small test networks. However, due to the size of the real decision problem, it is very difficult to solve it exactly, which is why a heuristic approach is required.
Within the scope of a master thesis, a heuristic procedure for solving the given optimization problem shall be developed, based on a literature survey. For this purpose, problem-specific and -adequate initial and improvement heuristics shall be designed and implemented, which allow to solve given problem instances based on real networks. Furthermore, the procedure shall be compared with other procedures on a number of given test networks with respect to solution quality and runtime. A basic understanding of mathematical optimization is required to work on this topic. Furthermore, basic programming skills (e.g. Python) are strongly recommended.
If you are interested, please contact Tjard Bätge