基于贝叶斯网络理论与仿生技术的翼形量水槽优化  

Optimization of Wing-shaped Measuring Flume Based on Bayesian Network Theory and Bionic Technology

作　　者：刘鸿涛 屈忠义[1] 李怡阳[2,3] 陈泽 LIU Hongtao;QU Zhongyi;LI Yiyang;CHEN Ze(Water Conservancy and Civil Eng.College,Inner Mongolia Agricultural Univ.,Hohhot 010018,China;School of Water Conservancy and Environmental Eng.,Changchun Inst.of Technol.,Changchun 130012,China;Jilin Province Water Eng.Safety and Disaster Prevention Eng.Lab.,Changchun 130012,China)

机构地区：[1]内蒙古农业大学水利与土木建筑工程学院,内蒙古呼和浩特010018 [2]长春工程学院水利与环境工程学院,吉林长春130012 [3]吉林省水工程安全与灾害防治工程实验室,吉林长春130012

出　　处：《工程科学与技术》2022年第6期105-115,共11页Advanced Engineering Sciences

基　　金：吉林省重点科技攻关项目(20170204008SF);吉林省高校科技与社科“十三五”科研规划项目(JJKH20200629KJ);吉林省科技厅重点科技研发项目(专项支持)(20180201036SF);国家自然科学基金项目(41761050)。

摘　　要：翼形量水槽的工作原理决定其工作必然会产生一定的水头损失,为保证量水槽自由出流的渠道比降随水头损失增大而增大,本文通过仿生优化减小翼形量水槽工作水头损失,扩大其在平原灌区渠道的适用范围。探索性地利用鸟类翅膀飞行减阻特性选取3种鸟翼作为仿生原型,截取仿生翼形曲线10条,按10个攻角水平应用处理得到78条仿生翼形量水槽外轮廓线;选取3个收缩比开展3种流量工况的仿真试验,采用极差分析法评价仿生翼形量水槽外轮廓线两因素的显著性水平和优水平;运用贝叶斯网络图模型推理分析方法解决仿生翼形曲线和攻角优组合的不确定性问题,并通过模型试验验证应用可行性。结果表明:翼形曲线是影响优化的主要因素,赛鸽鸽身与鸽翼交界截面曲线为优曲线、–10°为优攻角;攻角–5°与优曲线联合作用效果最佳,将优曲线按攻角–5°应用处理的仿赛鸽翼截面曲线作为优选仿生翼形量水槽外轮廓线;仿赛鸽翼截面曲线型量水槽各工况临界淹没出流状态为测流精度很高(测流平均误差约1.28%),壅水高度平均降低7.46%,水头损失平均降低5.81%,最大临界淹没度可达0.933,上游佛汝德数均小于0.4(可形成平稳缓流),最小工作比降可达1∶4 960。综上,利用仿生技术优化翼形量水槽可行,新型仿赛鸽翼截面曲线型量水槽综合量水性能很好,可用于灌区小比降渠道精细量水。Based on the working principle of a wing-shaped measuring flume,it will inevitably lead to head loss,and the channel gradient to ensure the free flow of measuring flume will increase with the increase of head loss.This paper proposes to reduce the working head loss of wing-shaped measuring flume by bionic optimization and expand its application scope in plain irrigation channels.In this paper,three kinds of bird wings are selected as bionic prototypes by exploring the drag reduction characteristics of bird wings.While 10 bionic wing curves are selected,78 bionic wing-shaped measuring flume contour lines are obtained according to 10 angles of attack.Three shrinkage ratios are selected to carry out simulation tests under three flow conditions.The range analysis method is used to evaluate the significance level and superior level of the two factors of bionic wing-shaped measuring flume contour lines.The Bayesian network diagram model reasoning analysis method is used to solve the uncertainty problem of the optimal combination of bionic wing-shaped curves and angles of attack,and the application feasibility is verified by model test. The results show that the wing-shaped curve is the main factor affecting the optimization, and the cross-section curve between the pigeon body and the pigeon wing is the optimal curve, and the optimal angle of attack is -10°;The combination of the angle of attack of -5° and the optimal curve has the best effect. The cross-section curve of the simulated racing pigeon wing treated by the optimal curve according to the angle of attack of -5° is used as the optimal contour line of the bionic wing-shaped measuring flume. The critical submerged outflow states of the new curved measuring flume with simulated racing pigeon wing section under various working conditions are as follows: the measuring accuracy is very high (the average error of measuring flow is about 1.28%);the backwater height is reduced by 7.46% on average;the head loss is reduced by 5.81% on average;the maximum critical subme

关 键 词：数值仿真 仿生 贝叶斯网络 极差分析 翼形量水槽 

分 类 号：S274.4[农业科学—农业水土工程]