arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.02585math.OC
收藏

最优控制问题中的L∞/L1对偶结果

L$\infty$/L1 Duality Results In Optimal Control Problems

Dan Goreac, Alain Rapaport

AI总结:

本研究提出最优控制问题的L∞/L1对偶结果,关联两类不同约束与代价形式的控制问题值函数,证明其在值函数下半连续初始条件下的对偶性,给出动力学正则性结构假设,并揭示非药物控制流行病相关研究的部分等价性。

摘要:

我们给出了一项对偶结果,将等周不等式G ≤ g_max约束下、具有上确界代价H的控制问题的值函数,与同一受控动力学下、具有代价G且状态约束为H ≤ h_max的控制问题的值函数关联起来。该对偶性在可保证值函数下半连续性的初始条件下得到证明,并结合最优性考量予以完善。此外,我们给出了可建立此类正则性的动力学结构假设。作为副产品,我们阐释了近期研究中,在峰值或预算限制下采用非药物控制手段的流行病相关研究之间的部分等价性。

英文摘要:

We provide a duality result linking the value function for a control problem with supremum cost H under an isoperimetric inequality G $\le$ gmax, and the value function for the same controlled dynamics with cost G and state constraint H $\le$ hmax. This duality is proven for initial conditions at which lower semi-continuity of the value functions can be guaranteed, and is completed with optimality considerations. Furthermore, we provide structural assumptions on the dynamics under which such regularity can be established. As a by-product, we illustrate the partial equivalence between recentworks dealing with non-pharmaceutically controlled epidemics under peak or budget restrictions.

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。

↑