摘要在我们实际的实验和勘探中，都会产生大量的数据。为了解释这些数据或者根据这些数据做出预测、判断，给决策者提供重要的依据。需要对测量数据进行拟合，寻找一个反映数据变化规律的函数。87693

本文介绍了几种常用的数据拟合方法，线性拟合、二次函数拟合、数据的n次多项式拟合等。然后重点介绍了非线性拟合的多元情况，推论了多元非线性数据拟合的通用数学模型，利用最小二乘法和极值原理，导出求解多元非线性回归方程的规范方程，并将其用矩阵进行表述，然后以毕业设计题目中所提出的问题为实例，对鱼类的数据进行了多元的非线性拟合，解决了鱼类数字特征间关系的疑问。随着计算机技术的发展，实验数据处理越来越方便。但也提出了新的课题，就是在选择数据处理方法时应该比以往更为慎重。因为稍有不慎，就会非常方便地根据正确的实验数据得出不确切的乃至错误的结论。所以提高拟合的准确度是非常有必要的。

毕业论文关键词：数据拟合   最小二乘法    曲线拟合    多元非线性拟合

毕业设计（论文）外文摘要

Title    Relationship between biological digital features 

Abstract

In our actual experiment and exploration, we will produce a lot of data。 In order to interpret the data or make predictions and judgments based on these data, it can provide important basis for decision makers。 Need to fit the measured data, looking for a function that reflects the change of the data。

This paper introduces several commonly used data fitting methods, linear fitting, two function fitting, data n degree polynomial fitting, etc。。 And then focuses on the nonlinear fitting of multiple conditions, deduces the general mathematical model of multivariate nonlinear data fitting, using the method of least squares and extreme value theory, derived multivariate nonlinear regression equation to the canonical equations and its matrix representation and the graduation design topic in the proposed problem as an example, the data on fish of multivariate nonlinear fitting, to solve the question of the relationship between digital characteristics of fish。 With the development of computer technology, the data processing is more and more convenient。 But also put forward a new topic, that is, in the choice of data processing methods should be more cautious than ever。 Because a little careless, it will be very convenient to draw the wrong conclusions based on the correct experimental data。 So it is very necessary to improve the accuracy of fitting。From+优!尔.YouErw.com 加QQ75201`8766

Keywords  data fitting, least squares, curve fitting, multivariate nonlinear fitting

目   次

1  绪论 1

2   数据拟合的方法简介 4

2。1 最小二乘法 4

2。2  线性拟合 5

2。3  二次曲线拟合 7

2。4  N次多项式拟合 9

2。5  用正交多项式系组成拟合函数的多项式拟合 11

2。6  指数函数拟合12

2。7  多元线性拟合 13

3 多元非线性拟合  15

3。1  数学模型 15

3。2 建立规范方程 16

4  实际运用   19

4。1  数据采集及处理 19

4。2  单个自变量的影响分析 20

4。3  假设拟合曲线解析式并用MATLAB进行拟合 23

4。4  数据检验 27