For the analysis and optimization of energy networks, HPA develops, implements and uses methods from numerical mathematics, data mining and machine learning. Examples for this are:

  • Development and implementation of solvers for differential-algebraic systems of equations (Flyer edaWorkshop 2011 [PDF, 2.2 MB])
  • Development of methods for the creation of interpolation and approximation models (response surfaces) und for statistics applications (DesParO)
  • Robust multiple criteria optimization: development and implementation of methods which are suitable and efficient for meta-modelling (DesParO)
  • Networks and graphs: HPA develops and implements a framework for network simulation (MYNTS für elektrical circuits, gas transport etc.) and software for network and graph manipulation and analysis (net'O'graph).

We also use proprietary, patented methods for statistical analysis and robust optimization of parameter-dependent chains of (simulation) processes (PRO-CHAIN).

For the highly scaled data analysis, both the algorithms for data processing as well as the data transport and their provision must fully exploit the possibilities of modern system architectures. Therefore, depending on requirements, methods of technical computing or cloud computing are also used.