摘要为实现我国世界第一造船大国的战略目标，控制造船费用的增长、提高经济效益成为了重要问题。针对横截面数据的船舶建造费用预算问题，由于主要参数比较相似的船舶装备其研制费用具有相似性，从而提出横截面数据的相似信息优先原理。86728

利用灰色系统建模软件分析船舶费用案例中各指标之间的相似度，区分横截面信息优先顺序，筛选变量。将不同年度造船费用之间的相似度进行归一化处理得出加权最小二乘法权重，建立相似信息优先的回归模型，并与其它两种方法进行比较，通过实例证明本文的方法精确度更高。相似信息优先的多元回归模型较好的反映相似信息之间的规律。论文网

在样本数据较少、变量间的线性相关程度较高的情况下，采用多元线性回归方程法会降低预算结果的有效性，从而引进GM(0，N)模型进行残差修正。用实例证明小样本数据也可以得出较高的精确度。

毕业论文关键词：相似信息原理；灰色系统理论；船舶建造费用；多元回归

Abstract

In order to achieve the strategic goals of the first nation of shipbuilding, the Chinese government made a lot of effort。 How to control the growth of the shipbuilding cost, improve the economic benefit has become an important problem。

The similar information priority principle is put forward in order to predict the ship construction cost from the data of the cross section, because the similar main parameters of ship construction cost lead to the similar cost。 The similarity between new researched ship and past ship is analyzed。 We put forward the prior principle of cross section data of similar information。 

We use the grey system modelling software to analyze the similarity among the indexes in the case, and distinguish between the priority of the cross section information。 Then, variables that we need are selected。 Through the similarity between different annual shipbuilding cost , we can obtain the weight from the normalized samples。 By establishing similar information priority of regression model, and comparing with other two methods, examples are presented to prove the higher accuracy of our method。 Similar information priority of multivariate regression model better reflects the laws between the similar information。

In cases of less data in the sample and high degree of linear correlation between variables, using multiple linear regression equation will reduce the validity of the budget results, thus GM (0, N) model is introduced to modify the residual error。 An example is used to prove that small sample data also can get higher accuracy。

Keywords: similar information priority principle；grey system theory；ship construction cost ；multiple regression model

                      目录

第一章 绪论 1

1。1 引言 1

1。2 方法 2

第二章 理论基础 3

2。1 建模步骤 3

2。2 相似优先原理 3

2。3 建立造船费用与参数的回归模型 5

2。4 GM(0，N)修正模型 6

第三章  船舶建造费用测算模型 7

3。1 背景 7

3。2 费用数据表示 8

3。3 相似度计算 9

3。3。1 不同年度造船费用的相似度 9