Improved dynamic programming for parallel reservoir system operation optimization

Xiang Zeng, Tiesong Hu, Ximing Cai, Yuliang Zhou, Xin Wang

Research output: Contribution to journalArticle

Abstract

Optimizing a multi-reservoir system is challenging due to the problem of the curse of dimensionality. In this paper, rule-based improved dynamic programming (RIDP) and stochastic dynamic programming (RISDP) algorithms for the optimal operation of a system with a number of parallel reservoirs are proposed to alleviate the dimensionality problem. The improvement is based on a key property: the monotonic dependence relationship between individual reservoir carryover storage and system water availability, which is derived with the assumption of the non-decreasing storage distribution characteristic of a parallel reservoir system. Furthermore, a diagnosis procedure is employed to remove infeasible state transitions, which enables the application of the monotonic relationship within the feasible solution space. In general, the computational complexity of (NS)n2 from DP can be reduced to (NS)n from RIDP (NS is the number of storage discretization for individual reservoirs, n is the number of reservoirs in a parallel system), with controlled solution accuracy. The improved algorithms are applied to a real-world parallel reservoir system in northeastern China. The results demonstrate the computational efficiency and effectiveness of RIDP and RISDP.

Original languageEnglish (US)
Article number103373
JournalAdvances in Water Resources
Volume131
DOIs
StatePublished - Sep 2019

Fingerprint

water availability
distribution

Keywords

  • Improved dynamic programming
  • Monotonic dependence relationship
  • Parallel reservoir system
  • State transitions

ASJC Scopus subject areas

  • Water Science and Technology

Cite this

Improved dynamic programming for parallel reservoir system operation optimization. / Zeng, Xiang; Hu, Tiesong; Cai, Ximing; Zhou, Yuliang; Wang, Xin.

In: Advances in Water Resources, Vol. 131, 103373, 09.2019.

Research output: Contribution to journalArticle

@article{b1b62a310888428298eed7c508c99c92,
title = "Improved dynamic programming for parallel reservoir system operation optimization",
abstract = "Optimizing a multi-reservoir system is challenging due to the problem of the curse of dimensionality. In this paper, rule-based improved dynamic programming (RIDP) and stochastic dynamic programming (RISDP) algorithms for the optimal operation of a system with a number of parallel reservoirs are proposed to alleviate the dimensionality problem. The improvement is based on a key property: the monotonic dependence relationship between individual reservoir carryover storage and system water availability, which is derived with the assumption of the non-decreasing storage distribution characteristic of a parallel reservoir system. Furthermore, a diagnosis procedure is employed to remove infeasible state transitions, which enables the application of the monotonic relationship within the feasible solution space. In general, the computational complexity of (NS)n2 from DP can be reduced to (NS)n from RIDP (NS is the number of storage discretization for individual reservoirs, n is the number of reservoirs in a parallel system), with controlled solution accuracy. The improved algorithms are applied to a real-world parallel reservoir system in northeastern China. The results demonstrate the computational efficiency and effectiveness of RIDP and RISDP.",
keywords = "Improved dynamic programming, Monotonic dependence relationship, Parallel reservoir system, State transitions",
author = "Xiang Zeng and Tiesong Hu and Ximing Cai and Yuliang Zhou and Xin Wang",
year = "2019",
month = "9",
doi = "10.1016/j.advwatres.2019.07.003",
language = "English (US)",
volume = "131",
journal = "Advances in Water Resources",
issn = "0309-1708",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Improved dynamic programming for parallel reservoir system operation optimization

AU - Zeng, Xiang

AU - Hu, Tiesong

AU - Cai, Ximing

AU - Zhou, Yuliang

AU - Wang, Xin

PY - 2019/9

Y1 - 2019/9

N2 - Optimizing a multi-reservoir system is challenging due to the problem of the curse of dimensionality. In this paper, rule-based improved dynamic programming (RIDP) and stochastic dynamic programming (RISDP) algorithms for the optimal operation of a system with a number of parallel reservoirs are proposed to alleviate the dimensionality problem. The improvement is based on a key property: the monotonic dependence relationship between individual reservoir carryover storage and system water availability, which is derived with the assumption of the non-decreasing storage distribution characteristic of a parallel reservoir system. Furthermore, a diagnosis procedure is employed to remove infeasible state transitions, which enables the application of the monotonic relationship within the feasible solution space. In general, the computational complexity of (NS)n2 from DP can be reduced to (NS)n from RIDP (NS is the number of storage discretization for individual reservoirs, n is the number of reservoirs in a parallel system), with controlled solution accuracy. The improved algorithms are applied to a real-world parallel reservoir system in northeastern China. The results demonstrate the computational efficiency and effectiveness of RIDP and RISDP.

AB - Optimizing a multi-reservoir system is challenging due to the problem of the curse of dimensionality. In this paper, rule-based improved dynamic programming (RIDP) and stochastic dynamic programming (RISDP) algorithms for the optimal operation of a system with a number of parallel reservoirs are proposed to alleviate the dimensionality problem. The improvement is based on a key property: the monotonic dependence relationship between individual reservoir carryover storage and system water availability, which is derived with the assumption of the non-decreasing storage distribution characteristic of a parallel reservoir system. Furthermore, a diagnosis procedure is employed to remove infeasible state transitions, which enables the application of the monotonic relationship within the feasible solution space. In general, the computational complexity of (NS)n2 from DP can be reduced to (NS)n from RIDP (NS is the number of storage discretization for individual reservoirs, n is the number of reservoirs in a parallel system), with controlled solution accuracy. The improved algorithms are applied to a real-world parallel reservoir system in northeastern China. The results demonstrate the computational efficiency and effectiveness of RIDP and RISDP.

KW - Improved dynamic programming

KW - Monotonic dependence relationship

KW - Parallel reservoir system

KW - State transitions

UR - http://www.scopus.com/inward/record.url?scp=85068836389&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068836389&partnerID=8YFLogxK

U2 - 10.1016/j.advwatres.2019.07.003

DO - 10.1016/j.advwatres.2019.07.003

M3 - Article

AN - SCOPUS:85068836389

VL - 131

JO - Advances in Water Resources

JF - Advances in Water Resources

SN - 0309-1708

M1 - 103373

ER -