Application of Linear Mean-Square Estimation in Ocean Engineering

作     者：王莉萍 陈柏宇 陈超 陈正寿 刘桂林 WANG Li-ping;CHEN Bai-yu;CHEN Chao;CHEN Zheng-shou;LIU Gui-lin

作者机构：College of Mathematical Science Ocean University of China Qingdao 266100 China College of Civil and Environmental Engmeermg University of California Berkeley Berkeley 94720 USA School of Naval Architecture and Marine Engineering Zhejiang Ocean University Zhoushan 316022 China College of Engineering Ocean University of China Qingdao 266100 China 

出 版 物：《China Ocean Engineering》 (中国海洋工程（英文版）)

年 卷 期：2016年第30卷第1期

页      面：149-160页

核心收录：

学科分类：081505[工学-港口、海岸及近海工程] 08[工学] 080401[工学-精密仪器及机械] 081105[工学-导航、制导与控制] 0804[工学-仪器科学与技术] 0815[工学-水利工程] 0824[工学-船舶与海洋工程] 0814[工学-土木工程] 0825[工学-航空宇航科学与技术] 082401[工学-船舶与海洋结构物设计制造] 0811[工学-控制科学与工程] 

基　　金：financially supported by the National Natural Science Foundation of China(Grant Nos.51379195 and 41476078) the Natural Science Foundation of Shandong Province(Grant No.ZR2013EEM034) the Scientific Research Foundation of Science Technology Department of Zhejiang Province(Grant No.2015C34013) the Science Research Program of Zhoushan(Grant No.2014C41003) the Innovation Fund for Graduate Student of Shandong Province(Grant No.SDYY12152) 

主　　题：ocean engineering linear mean square estimation N year return period storm surge 

摘      要：The attempt to obtain long-term observed data around some sea areas we concern is usually very hard or even impossible in practical offshore and ocean engineering situations. In this paper, by means of linear mean-square estimation method, a new way to extend short-term data to long-term ones is developed. The long-term data about concerning sea areas can be constructed via a series of long-term data obtained from neighbor oceanographic stations, through relevance analysis of different data series. It is effective to cover the insufficiency of time series prediction method＇s overdependence upon the length of data series, as well as the limitation of variable numbers adopted in multiple linear regression model. The storm surge data collected from three oceanographic stations located in Shandong Peninsula are taken as examples to analyze the number-selection effect of reference oceanographic stations（adjacent to the concerning sea area） and the correlation coefficients between sea sites which are selected for reference and for engineering projects construction respectively. By comparing the N-year return-period values which are calculated from observed raw data and processed data which are extended from finite data series by means of the linear mean-square estimation method, one can draw a conclusion that this method can give considerably good estimation in practical ocean engineering, in spite of different extreme value distributions about raw and processed data.