Monday, 10:30 - 10:55 h, Room: MA 550


Ali Koc
Parallel branch-cut-price for solving multistage stochastic unit commitment problems

Coauthor: Jayant Kalagnanam


Unit commitment (UC) lies in the heart of future smart grid. Independent system operators and utilities aim to solve various forms of the problem that handle such contemporary concepts as renewable generation, energy storage, power purchase agreements, future power markets, demand response, etc. These concepts induce various forms of uncertainty into the problem. We use multistage stochastic programming framework to incorporate these uncertainties, and present a parallel decomposition algorithm based on a branch-cut-price framework. We bring together several advancements in the UC literature, stochastic programming, mixed integer programming, large-scale optimization, and parallel computing. We develop a new weighting scheme and a lower bounding method to improve the decomposition algorithm, and a constructive heuristic method to restore near-optimal solutions. The serial algorithm solves problem instances as efficient as highly sophisticated commercial solvers, and the parallel algorithm solves large-scale nonlinear stochastic UC problem instances with up to 3000 generators, 24 hours, and 200 scenarios on a 32-processor cluster, obtaining almost linear speedups.


Talk 1 of the contributed session Mon.1.MA 550
"Unit commitment and inventory problems" [...]
Cluster 18
"Optimization in energy systems" [...]


  Payday Loans In Virginia. In this section we give only a brief summary recommendation for admission of Levitra. Full information can be found in the instructions for receiving medications with vardenafil.