Title :
A distributed optimization model for designing integrated operation policies of a multi-reservoir system
Author :
Tatano, H. ; Ruszcaynski, A.
Author_Institution :
Disaster Prevention Res. Inst., Kyoto Univ., Japan
Abstract :
An optimal operation rule design problem of a multiple-reservoir system is considered. The operation rule design problem is formulated as an expected loss per period minimization problem in infinite time horizon which is a problem of a Markov decision process (MDP). The paper presents a method to covert a large (centralized) multi-dimensional MDP problem to a sequence of problems which consist of a coordination LP problem and small decentralized MDP problems for each subsystem
Keywords :
Markov processes; decision theory; linear programming; minimisation; water supply; Markov decision process; distributed optimization model; expected loss per period minimization problem; infinite time horizon; integrated operation policies; multi-reservoir system; Costs; Design optimization; Equations; Information management; Portfolios; Queueing analysis; Reservoirs; Stochastic processes; Uncertainty; Water resources;
Conference_Titel :
Systems, Man, and Cybernetics, 1998. 1998 IEEE International Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-4778-1
DOI :
10.1109/ICSMC.1998.727620
