TU Darmstadt / ULB / TUprints

A Comprehensive Approach for an Approximative Integration of Nonlinear-Bivariate Functions in Mixed-Integer Linear Programming Models

Roth, Maximilian ; Franke, Georg ; Rinderknecht, Stephan (2022)
A Comprehensive Approach for an Approximative Integration of Nonlinear-Bivariate Functions in Mixed-Integer Linear Programming Models.
In: Mathematics, 2022, 10 (13)
doi: 10.26083/tuprints-00022093
Article, Secondary publication, Publisher's Version

[img] Text
mathematics-10-02226-v2.pdf
Copyright Information: CC BY 4.0 International - Creative Commons, Attribution.

Download (3MB)
Item Type: Article
Type of entry: Secondary publication
Title: A Comprehensive Approach for an Approximative Integration of Nonlinear-Bivariate Functions in Mixed-Integer Linear Programming Models
Language: English
Date: 2022
Place of Publication: Darmstadt
Year of primary publication: 2022
Publisher: MDPI
Journal or Publication Title: Mathematics
Volume of the journal: 10
Issue Number: 13
Collation: 17 Seiten
DOI: 10.26083/tuprints-00022093
Corresponding Links:
Origin: Secondary publication via sponsored Golden Open Access
Abstract:

As decentralized energy supply units, microgrids can make a decisive contribution to achieving climate targets. In this context, it is particularly important to determine the optimal size of the energy components contained in the microgrids and their optimal operating schedule. Hence, mathematical optimization methods are often used in association with such tasks. In particular, mixed-integer linear programming (MILP) has proven to be a useful tool. Due to the versatility of the different energetic components (e.g., storages, solar modules) and their special technical characteristics, linear relationships can often only inadequately describe the real processes. In order to take advantage of linear solution techniques but at the same time better represent these real-world processes, accurate and efficient approximation techniques need to be applied in system modeling. In particular, nonlinear-bivariate functions represent a major challenge, which is why this paper derives and implements a method that addresses this issue. The advantage of this method is that any bivariate mixed-integer nonlinear programming (MINLP) formulation can be transformed into a MILP formulation using this comprehensive method. For a performance comparison, a mixed-integer quadratic constrained programming (MIQCP) model—as an MINLP special case—is applied and transformed into a MILP, and the solution of the transformed problem is compared with the one of the MIQCP. Since there are good off-the-shelf solvers for MIQCP problems available, the comparison is conservative. The results for an exemplary microgrid sizing task show that the method delivers a strong performance, both in terms of approximation error (0.08%) and computation time. The method and its implementation can serve as a general user-tool but also as a basis for further methodological developments and research.

Uncontrolled Keywords: MILP; MINLP; MIQCP; big-M; sizing; scheduling; microgrids; linearization; nonlinear; bivariate
Status: Publisher's Version
URN: urn:nbn:de:tuda-tuprints-220931
Classification DDC: 600 Technology, medicine, applied sciences > 600 Technology
600 Technology, medicine, applied sciences > 620 Engineering and machine engineering
Divisions: 16 Department of Mechanical Engineering > Institute for Mechatronic Systems in Mechanical Engineering (IMS)
Date Deposited: 26 Aug 2022 12:10
Last Modified: 29 Aug 2022 07:45
URI: https://tuprints.ulb.tu-darmstadt.de/id/eprint/22093
PPN: 498686655
Export:
Actions (login required)
View Item View Item