"Plans are worthless, but planning is everything."
— Dwight Eisenhower
Uncertainty is the key reason for going from feedforward to feedback control. Yet, even very sophisticated modern control strategies such as MPC have to deal with uncertainty surrounding models, forecasts of exogenous signals (renewable generation, energy demand, and market prices). In principle one can distinguish two main approaches towards control and optimization under uncertainty:
- Worst-case considerations, which tend to be conservative
- Stochastic approaches, which require extra modelling efforts.
In this context, we work on stochastic optimization and control of nonlinear systems. Recent outcomes comprise toolboxes and new methods.
Selected publications:
- Ou, Ruchuan, Michael Heinrich Baumann, Lars Grüne, and Timm Faulwasser. "A Simulation Study on Turnpikes in Stochastic LQ Optimal Control." arXiv preprint arXiv:2010.12201 (2020).
- Mühlpfordt, Tillmann, Frederik Zahn, Veit Hagenmeyer, and Timm Faulwasser. "PolyChaos. jl--A Julia Package for Polynomial Chaos in Systems and Control." arXiv preprint arXiv:2004.03970 (2020).
- Appino, Riccardo Remo, Jorge Ángel González Ordiano, Ralf Mikut, Timm Faulwasser, and Veit Hagenmeyer. "On the use of probabilistic forecasts in scheduling of renewable energy sources coupled to storages." Applied energy 210 (2018): 1207-1218.