Page 107 - 2022年第53卷第5期

［10］ SHAHVERDI K，OMIDZADE F，GHODOUSI H . Comparing fuzzy SARSA learning and ant colony optimization
algor ithms in water delivery scheduling under water shortage conditions［J］. Journal of Irrigation and Drainage En⁃
gineering，2020，146（9）：04020028 .
［11］ SCHUURMANS J，BOSGRA OH，BROUWER R . Open-channel flow model approximation for controller design
［J］. Applied Mathematical Modelling，1995，19（9）：525-530 .
［12］ 苏海旺，管光华，钟乐，等 . 基于 ID 预测模型的明渠系统控制参数在线优化研究［J］. 中国农村水利水电，
2019（3）：141-144，151 .
［13］ HORVÁTH K，GALVIS E，VALENTÍN MG，et al . Is it better to use gate opening as control variable than dis⁃
charge to control irrigation canals?［J］. Journal of Irrigation & Drainage Engineering，2014，141（3）：04014054 .
［14］ SCHUURMANS J，CLEMMENS AJ . Modeling of irrigation and drainage canals for controller design［J］. Journal
of Irrigation & Drainage Engineering，1999，125（6）：338 .
［15］ LI Y . Offtake feedorward compensation for irrigation channels with distributed control［J］. IEEE Transactions on
Control Systems Technology，2014，22（5）：1991-1998 .
［16］ 吴怡，郑和震，雷晓辉，等 . 南水北调中线工程分水口敏感性研究［J］. 人民黄河，2018，40（7）：121-123，
128 .
［17］ 高学平，张岩，孙博闻，等 . 节制闸对分水口处渠道水力响应的影响及敏感性研究［J］. 水力发电学报，
2018，37（2）：79-87 .
［18］ 崔巍，李斯胜，陈文学，等 . 南水北调中线分水口群不同运用方式对总干渠水力控制特性的影响［J］. 水利
学报，2011，42（11）：1316-1321 .
［19］ 钟乐 . 基于参数辨识的渠道控制建模与仿真研究—以漳河灌区渠系为例［D］. 武汉：武汉大学，2016 .
［20］ CLEMMENS A，KACEREK T，GRAWITZ B，et al . Test cases for canal control algorithms［J］. Journal of Irriga⁃
tion and Drainage Engineering，1998，124（1）：23-30 .
［21］ 王长德，管光华 . 输水渠道系统运行仿真与控制软件：中国，2011SR034392［P］. 2011-03-01 .
［22］ HORVÁTH K . Model predictive control of resonance sensitive irrigation canals［D］. Catalunya：Universitat Po⁃
lite-cnica de Catalunya，2013 .
［23］ 管光华，钟锞，廖文俊，等 . 基于无量纲性能指标的渠系控制器参数优化［J］. 农业工程学报，2018，34
（7）：90-99 .
Modeling and validation based on Generalized-Integrator-Delay model for water conveyance
and distribution canal systems with multi-offtake

GUAN Guanghua， ZHU Zheli， WANG Kang
（School of Water Resources and Hydropower Engineering Science， Department of Agricultural Water Resources Engineering，
Wuhan University， Wuhan 430072， China）

Abstract：Integrator-Delay （ID） model is one of the most widely used control model for open canal sys⁃
tems，which assumes that all offtakes are located at the downstream end of the pools. However，the loca⁃
tion of the offtake may be anywhere along the canal for the irrigation districts in China. For such situation，
the application of ID model may result in poor control performance，or even the instability. Accordingly，
the paper first analyzes the impact of different location of offtakes on the prediction accuracy of ID model.
By considering the delay time of the disturbance caused by scheduled water orders，the Generalized-Integra⁃
tor-Delay （GID） model is proposed. Furthermore， model predict control （MPC） algorithm is implemented
on the basis of ID model and GID model，respectively. According to the simulation results，if the offtake
is located at the upstream in the pool，the proposed GID model has a higher accuracy for prediction，and
the improvement in control performance can reach 65%. During the controller tunning procedure，the option⁃
al domain of the controller parameters increases by 1.4 times，indicating that it is much easier to find an
appropriate controller，and the cost of trial-and-error can be greatly reduced.
Keywords：Integrator-Delay （ID） model；offtakes；model predict control
（责任编辑：李福田）

—607 —