摘 要

传统的人工组卷方法需要消耗大量的人力物力以及时间，这种方法在快速发展的现代化社会中已经远远满足不了人们的需求，于是新型高效的组卷方法的研究引起了人们的重视。本文在全面查阅参考文献和对组卷特点重点分析的基础上，设计并实现了基于遗传算法的智能组卷系统。

针对遗传算法组卷中存在的早收敛、容易陷入局部最优解等问题，本文从编码方式、选择算子、交叉算子以及变异算子等进行了相应的改进。采用分段实数编码的方式进行编码，缩短了编码的长度；采用带条件的初始化方法进行种群初始化，减少了适应度函数中的约束项；选择算子采用排序复选的方法，提高了优良个体的基因遗传率。本文采用C/S架构，以Visual Stduio2019为开发平台、.NET Framework为环境、C#为开发语言实现了智能组卷系统原型，我们实现了科目管理、试题管理、试卷查询以及智能组卷等功能模块。

通过系统测试表明：系统人机交互简单，采用改进的遗传算法，使用实数编码方式、排序复选的选择等技术手段,实现了高效且具有实用性的智能组卷,系统在可预见的各种情况下能满足设计要求。

关键词：智能组卷；遗传算法；分段实数编码；分段单点交叉

Abstract

The traditional manual method of generating test paper needs to consume a lot of manpower, material resources and time , which has been far from meeting people's needs in the rapidly developing modern society. Therefore, the research on the new and efficient method of generating test paper has attracted people's attention. In this paper, on the basis of a comprehensive review of references and analysis of the characteristics of test paper generation, an intelligent test paper generation system based on genetic algorithm is designed and implemented.

In order to solve the problem of premature convergence and easy to fall into the local optimal solution in generating test paper of genetic algorithm, some improvements have been made from coding method, selection operator, crossover operator and mutation operator. The method of segmented real number coding is used to shorten the length of coding; the method of conditional initialization is used to initialize the population, which reduces the constraints in fitness function; the method of sequence replication selection is used to select the operator, which improves the gene inheritance rate of excellent individuals. This paper uses C/S architecture, visual studio 2019 as the development platform, .Net framework as the environment and C# as the development language to realize the prototype of intelligent paper formation system. We have realized the functional modules such as subject management, test question management, examination paper query and intelligent test paper formation.

The system test shows that the system has simple human-computer interaction, using the improved genetic algorithm, using the real number coding method, the selection of sorting check and other technical means to achieve efficient and practical intelligent test paper formation, and the system can meet the design requirements under various foreseeable circumstances.

Key Words：Autogenerating Test Paper；Genetic Algorithm；Segmented Real Coding；Segmented Single Point Crossing

目 录

第1章 绪论 1

1.1 研究背景及意义 1

1.2 国内外研究现状 2

1.3 研究内容及目标 3

1.4 论文章节安排 3

1.5 本章小结 3

第2章 遗传算法组卷设计 4

2.1 试题属性与适应度函数 4

2.1.1 试题属性 4

2.1.2 适应度函数设计 4

2.2 初始化种群 6

2.2.1 分段实数编码 6

2.2.2 带条件初始化种群 7

2.3 选择算子 8

2.4 交叉算子 8

2.4.1 交叉概率与配对池 9

2.4.2 配对池内基于不相关系数的交叉 10

2.4.3 分段单点交叉 11

2.5 变异算子 11

2.6 组卷流程 13

2.7 本章小结 13

第3章 遗传算法组卷实现 14

3.1 带条件初始化种群 14