Reconstruction of missing boundary data for nonlinear parabolic systems using a control-type approach
摘要
This paper investigates the inverse Cauchy problem for nonlinear parabolic partial differential equations. We begin by studying the well-posedness of the associated boundary value problem (BVP). To address the intrinsic ill-posedness of the inverse problem, we reformulate it as a least-squares minimization problem, adding a vanishing regularization term. Numerical experiments conducted in both stationary and time-dependent regimes demonstrate the robustness of the proposed method to noise, its fast convergence, and its improved accuracy compared to the gradient descent method with a fixed step.