TY - JOUR
T1 - Convergence analysis of an asymptotic expansion regularization method for coefficient inverse problems in singularly perturbed time-periodic Burgers-type equations
AU - Chaikovskii, Dmitrii
AU - Zhang, Ye
AU - Liubavin, Aleksei
N1 - Publisher Copyright:
© 2026 Elsevier B.V.
PY - 2027/4
Y1 - 2027/4
N2 - This article investigates time-periodic singularly perturbed reaction-diffusion-advection equations with internal transition layers and the associated coefficient inverse problems. We consider two benchmark reconstruction settings: recovery of a spatial coefficient F(x,t)=f(x) from snapshot-type data and recovery of a temporal coefficient F(x,t)=f(t) in the time-periodic solution setting from boundary data and transition-layer positions measured over one period. In the latter formulation, measurements are taken on one full periodicity interval only. In the numerical implementation, the noisy data are periodically extended before applying the link equation, and the resulting pointwise coefficient values are then smoothed. No additional independent measurements are used outside the measured period. The main contribution is a rigorous convergence analysis of asymptotic-based inversion algorithms for these two settings. We derive explicit error estimates that separate the effects of the asymptotic approximation, spatial or temporal discretization, and measurement noise. The analysis is supported by a sharp estimate for the transition-layer location. Numerical experiments confirm the predicted convergence behavior and illustrate the stability of the proposed reconstructions under noisy data.
AB - This article investigates time-periodic singularly perturbed reaction-diffusion-advection equations with internal transition layers and the associated coefficient inverse problems. We consider two benchmark reconstruction settings: recovery of a spatial coefficient F(x,t)=f(x) from snapshot-type data and recovery of a temporal coefficient F(x,t)=f(t) in the time-periodic solution setting from boundary data and transition-layer positions measured over one period. In the latter formulation, measurements are taken on one full periodicity interval only. In the numerical implementation, the noisy data are periodically extended before applying the link equation, and the resulting pointwise coefficient values are then smoothed. No additional independent measurements are used outside the measured period. The main contribution is a rigorous convergence analysis of asymptotic-based inversion algorithms for these two settings. We derive explicit error estimates that separate the effects of the asymptotic approximation, spatial or temporal discretization, and measurement noise. The analysis is supported by a sharp estimate for the transition-layer location. Numerical experiments confirm the predicted convergence behavior and illustrate the stability of the proposed reconstructions under noisy data.
KW - Asymptotic expansion method
KW - Inverse problem
KW - Numerical methods
KW - Regularization
KW - Singularly perturbed partial differential equations
UR - https://www.scopus.com/pages/publications/105053570075
U2 - 10.1016/j.cam.2026.118248
DO - 10.1016/j.cam.2026.118248
M3 - Article
AN - SCOPUS:105053570075
SN - 0377-0427
VL - 494
JO - Journal of Computational and Applied Mathematics
JF - Journal of Computational and Applied Mathematics
M1 - 118248
ER -