Hi, I am Nils Löhndorf, and I am an Associate Professor at the Luxembourg Centre for Logistics and Supply Chain Management within the University of Luxembourg and Chairholder in Digital Procurement. As researcher and entrepreneur, I help decision-makers to make better in decisions in the face of uncertainty. My research interest are in stochastic optimization, in particular approximate dynamic programming, and its application to decision problems that involve uncertainty as they often occur in operations management.
Hydro Storage Optimization
In June of 2017, the Vienna University of Economics and Business featured my research on optimal operations and valuation of pumped-hydro storage power plants as part of their Researcher of the Month series. Check out the video, where I explain, why I believe that storing the energy of water is important and why optimizing it is worthwhile (in German):
Based on my research on stochastic optimization, I have developed a general-purpose solver for stochastic-dynamic optimization called QUASAR. The solver is intended for analysts, decision-makers, and researchers who want to solve difficult sequential decision problems that involve uncertainty. You can use QUASAR to solve linear multistage stochastic programs, continuous Markov decision processes, or stochastic-dynamic programs. QUASAR features
an algebraic modeling language for expressing continuous-state, finite-horizon, stochastic-dynamic decision problems.
a solution engine that combines scenario tree generation, approximate dynamic programming, and risk measures.
various functions and data structures to store, analyze, and visualize the optimal stochastic solution.
QUASAR is written in Java but provides native interfaces for Matlab and Python. If you want to try QUASAR go to quantego.com and sign up for a free trial. QUASAR is entirely free for academic usage.
Avila D, Papavasiliou A, Löhndorf N. 2021. Batch learning in stochastic dual dynamic programming. Available on Optimization Online. Download
Löhndorf N, Wozabal D. 2020. Gas storage valuation in incomplete markets. European Journal of Operational Research 288(1), 318-330. Preprint available on Optimization Online. Download
Löhndorf N, Shapiro A. 2019. Modeling time-dependent randomness in stochastic dual dynamic programming. European Journal of Operational Research 273(2), 650-661. Preprint available on Optimization Online. Download
Nersten S, Dimoski J, Fleten S-E, Löhndorf N. 2018. Hydropower reservoir management using a multi-factor price model with correlation between price and local inflow. In Proceedings of the 41st IAEE International Conference. International Association for Energy Economics.
Löhndorf N. 2016. An empirical analysis of scenario generation methods for stochastic optimization. European Journal of Operational Research 255(1), 121-132. Download
Löhndorf N, Riel M, Minner S. 2014. Simulation optimization for the stochastic economic lot scheduling problem with sequence-dependent setup times. International Journal of Production Economics 157, 170-176. Download
Löhndorf N, Wozabal D, Minner S. 2013. Optimizing trading decisions for hydro storage systems using approximate dual dynamic programming. Operations Research 61, 810-823. Download
Löhndorf N, Minner S. 2013. Simulation optimization for the stochastic economic lot scheduling problem. IIE Transactions 45, 796-810. Download
Transchel S, Minner S, Kallrath J, Löhndorf N, Eberhard U. 2011. A hybrid general lot-sizing and scheduling formulation for a production process with a two-stage product structure. International Journal of Production Research 49(9), 2463-2480. Download
Francas D, Löhndorf N, Minner S. 2011. Machine and labor flexibility in manufacturing networks. International Journal of Production Economics 131(1), 165-174. Download
Löhndorf N, Minner S. 2010. Optimal day-ahead bidding with renewable energies and storage. Energy Systems 1(1), 61-77. Download
Online Scenario Generator
Click here to generate scenarios from a multivariate distribution (normal, log-normal, uniform) for Monte Carlo simulation, numerical integration or stochastic programming. The generator is based on experiments with different scenario generation methods which are described in this article. The code is available on GitHub.