Optimal Distributed Lags and the Horizon Problem.
MASSACHUSETTS INST OF TECH CAMBRIDGE OPERATIONS RESEARCH CENTER
Pagination or Media Count:
The paper considers linear discrete control systems with general quadratic cost functions and partially observed states. It extends results due to Kalman on asymptotic stability of the optimal controls and as a by-product produces a computationally efficient method for obtaining first period linear decision rules for the problems treated for both finite and infinite horizons. Applications in production and economic planning are discussed and sufficient conditions on the nature of cost and system parameters are given under which first period decisions converge to a fixed value as the optimization time horizon increases. Finally, a characterization of the optimal asymptotic controls is given for the discounted and undiscounted cases. Author
- Economics and Cost Analysis
- Statistics and Probability
- Operations Research