Dynamic Programming Principle and Viscosity Solutions of Hamilton–Jacobi–Bellman Equations for Stochastic Recursive Control Problem with Non-Lipschitz Generator
- 38 Downloads
In this paper, we consider the stochastic recursive control problem under non-Lipschitz framework. More precisely, we assume that the generator of the backward stochastic differential equation that describes the cost functional is monotonic with respect to the first unknown variable and uniformly continuous in the second unknown variable. A dynamic programming principle is established by making use of a Girsanov transformation argument and the BSDE methods. The value function is then shown to be the unique viscosity solution of the associated Hamilton–Jacobi–Bellman equation via truncation methods, approximation techniques and the stability result of viscosity solutions.
KeywordsStochastic recursive control problem Non-Lipschitz generator Hamilton–Jacobi–Bellman equation Viscosity solution
Y. Zhuo is supported by the National Natural Science Foundation of China (No. 11171076), and by Science and Technology Commission, Shanghai Municipality (No. 14XD1400400). Y. Dong is supported by Région Pays de la Loire throught the grant PANORisk. J. Pu is supported by the National Natural Science Foundation of China (No. 11701371), and by XuLun Scholar Project of Shanghai Lixin University of Accounting and Finance.
- 17.Liu, Z.: Control and differential game of forward-backward stochastic systems. PhD thesis, Fudan University (2016)Google Scholar
- 19.Pardoux, E.: BSDEs, weak convergence and homogenization of semilinear PDEs. In: Clarke, F.H., Stern, R.J. (eds.) Nonlinear analysis, differential equations and control, pp. 503–549. Kluwer Academic, Dordrecht (1999)Google Scholar
- 23.Yan, J.A., Peng, S., Fang, S.Z., Wu, L.M.: Several topics in stochastic analysis. Academic Press of China, Beijing (1997)Google Scholar