基于贝叶斯网络钱塘江河口地区咸潮入侵研究.pdfVIP

Journal of Water Resources Research 水资源研究, 2012, 1, 475-479 doi:10.4236/jwrr.2012.16076 Published Online December 2012 (/journal/jwrr.html) Analysis on Saline Water Intrusion in Qiantang River * Based on Bayesian Network Xichao Gao, Chong Ma, Xujie Zhang, Yueping Xu, Yongqiang Zhang Institute of Hydrology and Water Resources, Civil Engineering, Zhejiang University, Hangzhou Email: pandagxc@ Received: Jul. 12th st th , 2012; revised: Aug. 1 , 2012; accepted: Aug. 14 , 2012 Abstract: Saline water intrusion in Qiantang River exerts a severe impact on residential life, environment and economic development. In this article, a simple Bayesian Network model based on water levels of Tong- lu, Zhakou and Kanpu stations is established and the parameters are analyzed through the observation data. Whether the saline water intrusion will occur or not is determined by the relative water level of Tonglu, Zha- kou and Kanpu stations. Based on the model, the probability of saline water intrusion at Tonglu and Zhakou stations is determined. The reasoning ability of Bayesian Network is utilized to simulate the potential saline water intrusion in some certain situations. Keywords: Bayesian Network; Saline Water Intrusion; Estuary of Qiantang River; Joint Risk 基于贝叶斯网络的钱塘江河口地区咸潮入侵研究* 高希超，马 冲，张徐杰，许月萍，张永强 浙江大学建筑工程学院水利工程学系，杭州 Email: pandagxc@ 收稿日期：2012 年 7 月 12 日；修回日期：2012 年 8 月 1 日；录用日期：2012 年 8 月 14 日 摘 要：本文建立了钱塘江下游桐庐、闸口及瞰浦三站水(潮)位之间的贝叶斯网络模型，并对网络模 型的参数进行了估计和分析。通过三站高、低水位的不同遭遇情况来确定是否有咸潮入侵的情况发生， 给出了桐庐、闸口两站咸潮入侵的概率，且通过贝叶斯网络的推理功能对所研究地区的咸潮入