为了模拟黄河内蒙古段冰情的发生、发展过程，以可变模糊集理论为基础，提出了可变模糊聚类循环迭代模型来预报黄河内蒙古段巴彦高勒站流凌日期。选定初始权重矩阵、初始模糊聚类矩阵、初始模糊聚类中心矩阵，应用迭代公式求得最优权重矩阵、最优模糊聚类矩阵、最优模糊聚类中心矩阵，建立符合相关性要求的回归方程，从而实现预报识别检验。将模型预报结果与实测值对比分析，预报结果合格率达到80%，属于甲等预报方案，表明基于物理成因分析和数学理论的可变模糊聚类循环迭代模型可用于黄河内蒙古段的流凌预报。% In order to simulate the initiation and development of ice regimes in the Inner Mongolia reach of Yellow River, a variable fuzzy clustering loop iteration model is proposed to forecast the ice-run date at Bayangaole station in the Inner Mongolia Reach of Yellow River based on the variable fuzzy set theory. The matrices of the initial weight, the initial fuzzy clustering and the initial fuzzy clustering center are chosen, and their optimal matrices are obtained by means of the loop iteration formulae. Regression equations to satisfy the correlation are established to achieve forecast examination. The forecast results are compared with the measured values, and the acceptability is 80%. The forecast schemes belong to the first-class ones. It indicates that the proposed variable fuzzy clustering loop iteration model based on physical factors and mathematical theories can be employed to predicate the ice-run date in the Inner Mongolia reach of Yellow River.
展开▼
