黄河断流严重程度分级与判别方法

Classification and Identification of Severity Degree of the No-flow Events in the Yellow River

根据１９７２年以来有关黄河断流的观测数据，选择与断流特点有关的特征因子组合，给出了能够反映断流状况的综合指标．并据此结合聚类分析得出了关于黄河断流严重程度的分类结果，指出了相应的分级原则，基于断流级别与河流径流量及其年内径流变差系数之间可能存在的密切关系。建立了反映三者关系的人工神经网络模型。用以判别已知河流来水特性条件下的黄河下游断流严重程度。此方法为定量描述和预测黄河断流的程度奠定了基础。

Based on a review of the characteristic factors used for no-flow behavior analysis in previous studies, this paper proposed an additional comprehensive index which could remedy some deficiencies of other factors and thereby constitute a basis for classification of the severity degree of the Yellow River no-flow events. According to the cluster analysis of the measured data since 1972 when the no-flow phenomena occurred in the Yellow River first time, five severity levels could be classified for no-flow events in the Yellow River. Furthermore, the severity level is closely related to the incoming runoff as well as its fluctuation within a year expressed in terms of a deviation coefficient. A BP-typed ANN model was developed for identifying the complex response of the severity level to flow characteristics such as the incoming runoff and its fluctuation. The proposed method is of significance to the quantitative assessment of the severity level of the no-flow events in the Yellow River, and its applicability was proved through the comparison between the predicted and the measured results.