A robust numerical method for a control problem involving singularly perturbed equations

详细信息    查看全文

• 作者：Alejandro Allendes ^{alejandro.allendes@usm.cl" class="auth_mail" title="E-mail the corresponding author} ; Erwin Herná ; ndez ^{erwin.hernandez@usm.cl" class="auth_mail" title="E-mail the corresponding author} ; Enrique Otá ; rola ; ^{enrique.otarola@usm.cl" class="auth_mail" title="E-mail the corresponding author}
• 关键词：Linear&ndash ; quadratic optimal control problem ; Convection&ndash ; reaction&ndash ; diffusion equation ; Singularly perturbed equation ; Finite elements ; Graded meshes ; Anisotropic estimates
• 刊名：Computers & Mathematics with Applications
• 出版年：2016
• 出版时间：August 2016
• 年：2016
• 卷：72
• 期：4
• 页码：974-991
• 全文大小：1418 K

文摘

We consider an unconstrained linear–quadratic optimal control problem governed by a singularly perturbed convection–reaction–diffusion equation. We discretize the optimality system by using standard piecewise bilinear finite elements on the graded meshes introduced by Durán and Lombardi in (Duŕan and Lombardi 2005, 2006). We prove convergence of this scheme. In addition, when the state equation is a singularly perturbed reaction–diffusion equation, we derive quasi-optimal a priori error estimates for the approximation error of the optimal variables on anisotropic meshes. We present several numerical experiments when the state equation is both a reaction–diffusion and a convection–reaction–diffusion equation. These numerical experiments reveal a competitive performance of the proposed solution technique.