arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2401.10583math.OC
收藏

由稳态拟线性偏微分方程控制的最优控制的存在性与松弛

Existence and relaxation for optimal control governed by steady, quasilinear PDEs

Pablo Pedregal

AI总结:

本文研究椭圆型拟线性 PDE 约束的最优控制,分别分析变分状态方程、非变分单调算子及含散度外非线性项三种情形,证明前两类最优解存在并为第三类建立松弛框架。

摘要:

我们关注由椭圆型拟线性偏微分方程(PDE)控制的最优控制问题。尽管文献中已有此类问题的多种例子,但我们尝试通过处理三种基本情形来描述一些一般原理。在第一种情形中,我们假设状态方程是变分的;第二种情形聚焦于作为状态方程的非变分单调算子;最后,我们在方程的散度部分之外加入一个非线性项。在前两种情形中,最优解的存在性可在适当的假设集合下建立,而第三种情形则需要进行松弛。关于代价泛函,尽管可以处理更一般的例子,但我们将采用一个由两项组成的典型情形:一项依赖于状态,另一项具有典型 Tychonov 正则化的形式。

英文摘要:

We focus on optimal control problems governed by elliptic, quasilinear PDEs. Though there are various examples of such problems in the literature, we make an attempt at describing some general principles by dealing with three basic situations. In the first one, we assume that the state equation is variational; the second one focuses on a non-variational, monotone operator as state equation; finally, we add a non-linear term off the divergence part of the equation. In the first two cases, existence of optimal solutions can be established under suitable sets of assumptions, while relaxation is required for the third situation. Concerning the cost functional, and though more general examples can be dealt with, we will take a typical case consisting of two terms: one depending on the state, and another one of the form of a typical Thychonov regularization.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑