Solar towers use many mirrors to concentrate sun light on a central, tower-mounted receiver. The receiver then transfers the resulting heat to a fluid (i.e. molten salt or air) that, in turn, exchanges the heat to steam which powers a turbine, generating electricity. The placement of the mirrors may lead to individual mirrors being blocked and shaded; this affects the efficiency (and therefore costs) of the power plant. The model is later used for an optimization process which finds the most efficient arrangement of mirrors.
Within this practical course several problems are offered, each will be solved by a group of two to three students.
- Validation Validate the existing model by writing Gtests for all sub-models. Validate against third-party software.
- Annual Simulation Accelerate the annual simulation with a smart choice of time points in a year. Use DistMesh and RBF.
- Optimal Cabling Find an optimal path for laying the cables which connect heliostats in a solar tower power plant. Alternatively consider costs for an autonomous system (PV, battery, WiFi).
- Optimization Interface Accelerate the interface between model and optimizer class
- Genetic Algorithm Use a GA library to allow optimizations with the use of a genetic algorithm.
- Optimizer for Groups Develop an optimizer for heliostat groups.
- Extended Optimizer Extend the optimizer in such a way that different heliostat types can be considered, and the optimal number of heliostats can be found (depending on the objective function.
- Test Optimizers Set up test cases for existing solar tower power plants as Helio100, PS10, Hami and Ashalim. Test all existing optimizer incl. the local search and Multi-Step Optimizer.
For all topics basic skills in C++ is needed. Please state in which topic you are interested the most. If you already have a favorite partner, please indicate him/her, such that we could select you both during the registration process.
This project is a corporation of the Theory of Hybrid Systems (i2) research group headed by Prof. Dr. Erika Ábrahám, and the research group for Continuous Optimization at IGPM. The project will be co-supervised by Pascal Richter. For further questions please contact Pascal Richter.