Mixed-integer nonlinear optimization methods for the optimization of the overall system Clean Circles under uncertainty

The aim of the project is to algorithmically optimize the coupled material and energy flows in the cycles with regard to one or more target variables under appropriate constraints on the basis of the overall model developed. Since the system components are generally described by non-linear algebraic or differential algebraic equations and discrete decisions have to be made when configuring the overall system (combination of different technologies, selection of scenarios, etc.), these are mixed-integer non-linear optimization problems. In addition, uncertainties must be taken into account both in the modelling of the iron cycle and in the inclusion of market aspects.

Project aims

  • Development of globally solvable mixed-integer nonlinear optimization problems for the Clean Circles system with focus on the technical component
    • Features: nonlinear and differential-algebraic constraints (model equations), discrete decisions (technology scenario)
  • Development of efficient solvers. Particular challenges: efficient handling of nonlinear and differential-algebraic components
  • Handling of uncertainties by techniques of nonlinear robust optimization
  • Multi-criteria optimization by sensitivity-based techniques to explore the Pareto front
  • Coupling / combination with optimization model for logistic network (project M. Pfetsch)
  • Optimization techniques to learn (grey box-) models from data and integration in the optimization problem
  • Application to realistic instances of the Clean Circles system

Team

Technical University of Darmstadt

Contact

work +49 6151 16-23430

Work Schlossgartenstr. 7
64289 Darmstadt

Technical University of Darmstadt

Contact

work +49 6151 16-23443

Work S4|10 17
Dolivostr. 15
64293 Darmstadt