海底冲淤变化预测的聚类指数增长模型研究

Research of Cluster Index Growth Model to Predict Submarine Erosion and Sediment

提出了一种新的聚类指数增长模型,并用该模型预测了江苏洋口港海区西太阳沙附近海底的冲淤变化。首先采用聚类分析法对空间连续的水深冲淤变化数据序列进行分类,然后针对各类具有不同的冲淤变化趋势的区域分别建立指数增长预测模型。经比较分析预测结果,发现在分类数较大及预测时间不太长的情况下,能够取得满意的预测结果;但随预测时间的增加,要不断的修正参数值,才能得到切合实际的预测结果,所以该模型还具有一定的时间局限性,还有待于改进。该模型的建立不但可以为江苏省东部沿海海港建设提供科学的决策依据,也可以成为近岸海底冲淤变化预测的发展方向之一,有较高的理论意义和广远的应用前景。

A new cluster index growth model to predict submarine erosion and sediment near the West Tai-Yang-Sha of Jiangsu Yangkou-Port has been brought forward in this paper. The first step is to classify the continuous spatial depth data sequence according to the original data by clustering. Then, a group of index growth models with their respective transformation tendency should be established. After comparing and analyzing the predicted results, we can find that if the number of categories is large and the span of prediction is not very long, it can generate satisfactory results; but as the time span for prediction increases, the parameter must be rectified gradually to get the appropriate predictive result. Hence, this model has time limitation and needs improvement. To summarize, this model offers scientific basis of decision-making for seaport construction in the south of Jiangsu Province. Meanwhile, it can become a new direction of development to predict the submarine erosion and sediment of seashore. Consequently, this model has a profound theoretical meaning and wide application prospect.