Efficient numerical treatment of aggregation integrals in multivariate population balance equations

Publikationstyp
Thesis
Thesis Type
Doctoral Thesis
Publikationsdatum
2020
Sprache
English
Author
Ahrens, Robin 
Advisor
Le Borne, Sabine orcid-logo
Referee
Benner, Peter 
Title Granting Institution
Technische Universität Hamburg
Place of Title Granting Institution
Hamburg
Examination Date
2020-06-08
Institut
Mathematik E-10 
DOI
10.15480/882.2851
TORE-URI
http://hdl.handle.net/11420/6907
Lizenz
http://rightsstatements.org/vocab/InC/1.0/
Citation
Technische Universität Hamburg (2020)
This work develops efficient numerical methods for aggregation integrals in multivariate population balance equations on a uniform tensor grid. These are based on the fast Fourier transform and the tensor-train format for the efficient storage of the resulting data structure. The inverse problem of kernelestimation from discrete in time data is additionally addressed. All results are underlined with numerical simulations.
Schlagworte
Population balance equation
Fast Fourier transformation
Tensor decomposition
Kernel estimates
Moment conservation
Multivariate Convolution
DDC Class
500: Naturwissenschaften
510: Mathematik