Title :
The optimum stochastic control for a problem with random cost incursion times
Author_Institution :
Analysis and Technology, Incorporated, North Stonington, CT, USA
fDate :
10/1/1986 12:00:00 AM
Abstract :
A sequential stochastic control problem is considered in which the performance criterion cost is incremented periodically at random times. Although these cost incursion times are unknown, they are governed by a known prior probability density function. The optimum Bayesian control decision strategy is obtained by solving the stochastic dynamic programming equation. As anticipated, the solution exhibits a type of separation principle.
Keywords :
Bayes procedures; Discrete-time systems; Dynamic programming; Stochastic optimal control, linear systems; Bayesian methods; Cost function; Data mining; Dynamic programming; Equations; Monitoring; Observers; Probability density function; Stochastic processes; Vectors;
Journal_Title :
Automatic Control, IEEE Transactions on
DOI :
10.1109/TAC.1986.1104135