用模式分类方法区分高、低含沙水流  

Distinguish hyper-concentration flow from lower-concentration flow by mode clarify method

作　　者：陈雄波[1] 唐洪武[2] 丁大发[1] 

机构地区：[1]黄河勘测设计有限公司,河南郑州450003 [2]河海大学水利水电工程学院,江苏南京210098

出　　处：《水科学进展》2005年第4期500-505,共6页Advances in Water Science

摘　　要：在介绍模式分类方法的基础上,通过分析BP神经网络训练结果发现:高、低含沙水流挟沙力的变化规律有相似之处,说明挟沙规律不宜作为区别高、低含沙水流的标准,与前人的分析比较吻合;当多沙流体宾汉剪应力大于或小于某一数值组次的资料占总数的绝大多数时,这些组次训练结果与实测值符合较好,而另外组次训练结果与实测值有较大差别,这说明低含沙流体随着含沙量的增加流变方程由牛顿体向宾汉体模型转化时对应的宾汉剪应力有一个临界值。因此可将多沙流体的宾汉剪应力是否达到相应的临界值,作为区分高、低含沙水流的标准。对电木粉高含量的流体,该临界值约为3.2×10-1Pa。The theoretical foundation of mode clarification is introduced at first, based on which the analysis is done according to the results got by the Back-Propagation neural (BPN) new training. It is found that the features of the sediment capacity ability are similar to the hyper-concentration and lower-concentration flow. So the sediment capacity ability is not a desirable criterion for the classification of hyper-concentration and lower-concentration flow. This agrees well with the results got by other researchers. During the BPN training, if, in most cases, the Bingham shear stress of the flow is greater or less than certain critical value, the training results always agree well with the measured data, while others not. Several training tests get nearly the same conclusions. It can be deduced that there indeed exists a critical value of the Bingham shear stress, and when the shear stress exceeds this critical value, the flow can be classified as the hyper-concentration flow. For the flow with a hyper-concentration powered phenolic, this critical value is about 3.2×10-1 Pa.

关 键 词：高含沙水流 模式分类 人工神经网络 挟沙力 宾汉剪应力 

分 类 号：TV143[水利工程—水力学及河流动力学]