Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Research › peer-review
Identifiability Analysis for Source Problem of Quasi-Hyperbolic Equation. / Zvonareva, Tatiana; Krivorotko, Olga.
2023 5th International Conference on Problems of Cybernetics and Informatics, PCI 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 1-4 (2023 5th International Conference on Problems of Cybernetics and Informatics, PCI 2023).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Research › peer-review
}
TY - GEN
T1 - Identifiability Analysis for Source Problem of Quasi-Hyperbolic Equation
AU - Zvonareva, Tatiana
AU - Krivorotko, Olga
N1 - Conference code: 5
PY - 2023
Y1 - 2023
N2 - The source for a quasi-hyperbolic equation with a small parameter before the second derivative in time using additional measurements of integral type in fixed times is investigated. The source is parametrized by 6 constants. A sensitivity-based identifiability analysis of the source problem is carried out using the Sobol method. It is shown that all investigated source parameters are not enough sensitive to the additional measurements. The source problem has been reduced to a misfit function minimization problem and solved by the tensor train global optimization method. For 6 parameters it is shown that the smallest error value of the reconstruction of the required parameters is achieved in the case of non-zero small parameter. The reducing of the number of parameters to 3 is a regularization.
AB - The source for a quasi-hyperbolic equation with a small parameter before the second derivative in time using additional measurements of integral type in fixed times is investigated. The source is parametrized by 6 constants. A sensitivity-based identifiability analysis of the source problem is carried out using the Sobol method. It is shown that all investigated source parameters are not enough sensitive to the additional measurements. The source problem has been reduced to a misfit function minimization problem and solved by the tensor train global optimization method. For 6 parameters it is shown that the smallest error value of the reconstruction of the required parameters is achieved in the case of non-zero small parameter. The reducing of the number of parameters to 3 is a regularization.
KW - diffusion-logistic model
KW - identifiability
KW - inverse problem
KW - optimization
KW - regularization
KW - sensitivity analysis
KW - source problem
UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85179896046&origin=inward&txGid=c0c5c82942fdbc1cd75e4f2a7bfbf225
UR - https://www.mendeley.com/catalogue/d6858743-b6c1-3ee4-91db-b3a473dbb5d0/
U2 - 10.1109/PCI60110.2023.10325964
DO - 10.1109/PCI60110.2023.10325964
M3 - Conference contribution
SN - 9798350319064
T3 - 2023 5th International Conference on Problems of Cybernetics and Informatics, PCI 2023
SP - 1
EP - 4
BT - 2023 5th International Conference on Problems of Cybernetics and Informatics, PCI 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 5th International Conference on Problems of Cybernetics and Informatics
Y2 - 28 August 2023 through 30 August 2023
ER -
ID: 59681718