加载中…
In this note, we give the stochastic maximum principle for optimal control of stochastic PDEs in the general case (when the control domain need not be convex and the diffusion coefficient can contain a control variable).

生成讲解博客的中文版本

生成AI智能校验的中文版本