The generalized principal eigenvalue for Hamilton-Jacobi-Bellman equations of ergodic type (Q2349405)

The paper is concerned with the Hamilton-Jacobi-Bellman (HJB) equation of ergodic type. The unknowns are the eigenvalue and the eigenfunction of a certain nonlinear additive eigenvalue problem, called the ergodic problem. The ergodic problem is closely related to the stochastic control problem. The objective of this paper is to investigate several qualitative properties of the generalized principal eigenvalue problem. The paper consists of two parts. In the first part, the author explores the relationship between the generalized principal eigenvalue and the optimal value of the minimizing problem for the stochastic control. More specifically, the author discusses a necessary and sufficient condition for the two values to coincide. In the second part, the author discusses more specific (qualitative) properties of the generalized principal eigenvalue problem. In particular, the mapping from the parameter of the ergodic problem to the principal eigenvalue is studied under some restrictive assumptions on the coefficients of the ergodic problem. The paper continues the previous work of the same author, e.g., the work published in [J. Math. Pures Appl. (9) 100, No. 3, 368--390 (2013; Zbl 1295.35092)].