Burlacu R, Egger H, Gross M, Martin A, Pfetsch ME, Schewe L, Sirvent M, Skutella M (2017)
Publication Language: English
Publication Type: Other publication type
Publication year: 2017
URI: https://opus4.kobv.de/opus4-trr154/frontdoor/index/index/docId/221
We consider the optimization of a gas network that is described by a coupled system of parabolic partial differential equations. Energy and mass balance at network junctions and active elements, like compressors or valves, are modeled as additional algebraic equations. The resulting optimal control problem is discretized in space and time by a particular finite volume method, which can be shown to be well-posed under rather general assumptions. This discretize-first-then-optimize
procedure yields a mixed-integer nonlinear problem (MINLP) that can be solved to global optimality.
For the numerical solution of the MINLP, we consider a relaxation approach allowing to solve the problem globally by a sequence of mixed-integer problems (MIPs) with any required accuracy. The relaxation is based on piecewise linearization of the nonlinear constraints modeling the gas dynamics on the pipelines. Due to the particular discretization of the state equations, only univariate nonlinearities have to be approximated. This substantially facilitates the numerical treatment of the nonlinear constraints. To illustrate the efficiency of the proposed approach, we present numerical tests for typical benchmark problems.
APA:
Burlacu, R., Egger, H., Gross, M., Martin, A., Pfetsch, M.E., Schewe, L.,... Skutella, M. (2017). A Global Optimization Approach for Instationary Gas Transport in Pipeline Networks.
MLA:
Burlacu, Robert, et al. A Global Optimization Approach for Instationary Gas Transport in Pipeline Networks. 2017.
BibTeX: Download