文章摘要
刘晋龙,王忠静,杨志刚,张腾,朱现坡.灌区输配水分段积分时滞模型及自适应预测控制方法[J].水利学报,2023,54(2):232-243
灌区输配水分段积分时滞模型及自适应预测控制方法
Segment integrator delay model and adaptive predictive control method for water transmission and distribution system in irrigation districts
投稿时间:2022-01-27  
DOI:10.13243/j.cnki.slxb.20220068
中文关键词: 水联网  灌区  输配水系统  渠系自动控制  分段积分时滞模型  自适应控制
英文关键词: internet-of-water  irrigation district  water transmission and distribution system  automatic canal control  segment integrator delay model  adaptive control
基金项目:宁夏重点研发计划项目(2020BCF01002);国家重点研发计划项目(2021YFD1900600);青海省重大科技专项(2019-SF-A4)
作者单位E-mail
刘晋龙 清华大学 水利水电工程系, 北京 100084
北京市市政工程设计研究总院有限公司, 北京 100082 
 
王忠静 清华大学 水利水电工程系, 北京 100084
清华大学 水沙科学与水利水电工程国家重点实验室, 北京 100084
宁夏大学 西北土地退化与生态恢复省部共建国家重点实验室培育基地, 宁夏 银川 750021 
zj.wang@tsinghua.edu.cn 
杨志刚 清华大学 水利水电工程系, 北京 100084  
张腾 清华大学 水利水电工程系, 北京 100084  
朱现坡 清华大学 水利水电工程系, 北京 100084  
摘要点击次数: 2019
全文下载次数: 921
中文摘要:
      大型灌区渠道长、流量大、取水口多且分散,积分时滞模型(ID)适用性不强、参数不易率定;以水联网灌区的实时观测体系为基础,研究各取水口流量变化传播的时滞差异、非回水区蓄水量变化对回水区水位的影响,提出分段积分时滞模型(SID),建立在线自适应预测控制(APC)算法,在宁夏西干渠百万亩灌区调试验证。结果显示,基于ID的线性二次型控制(LQ)和模型预测控制(MPC)在水位目标不变时,可稳定水位,但超调量大,在水位目标显著增加时,控制失效;基于SID的模型预测控制(MPC)和自适应预测控制(APC)可将超调量控制在较小范围以内,且APC误差更小、控制效果更优。
英文摘要:
      In large irrigation district, canal pools have large length and flow rate, with many scattered water offtake gates, Integrator Delay (ID) model is not very applicable and its parameters are not easy to rate.Based on the real-time observation system of internet-of-water irrigation district, segmenting single canal pool, considering the time delay differences of each offtake gate and the effect of non-backwater area storage changes on the water level in backwater area, Segment Integrator Delay (SID) model is proposed, and Adaptive Predictive Control (APC) algorithm with online identification is established, the performance of the algorithm is verified by simulation in Ningxia Xiganqu Canal.The results show that the ID-based Linear Quadratic control (LQ) and Model Predictive Control (MPC) can stabilize the water level with constant water level target, but the overshoot is large, control essentially fails when the water level target increases significantly.SID-based Model Predictive Control (MPC) and Adaptive Predictive Control (APC) can control the overshoot to within small range, moreover, APC provides better control due to smaller system model errors.
查看全文   查看/发表评论  下载PDF阅读器
关闭