摘 要

海洋中蕴含了丰富的资源，能够补充当今陆地匮乏的能源，所以人们开始了对海洋的深度开发。随着人们探索海洋的过程中对智能型水下机器人的需求不断提高， AUV智能化的水平也应该不断提高，因此AUV的运动控制系统及路径规划显得尤为重要。

本文主要研究了AUV的路径规划问题。在本文的第二章节中，首先对AUV路径规划问题进行简要的描述分析，然后介绍了几种常见的AUV路径规划的算法，并且介绍了对应算法的国内外研究成果。然后对本文所研究的遗传算法包括其求解步骤进行详细阐述，由于遗传算法在求解过程中会出现局部最优的情况，不再继续寻找最优解，所以在本文中提出了在遗传算法中引入模拟退火算法的思想，改进遗传算法得到遗传模拟退火算法。第三章中首先介绍了五种环境建模的常用方法，分别为可视图法、Voronoi法、栅格法、单元树法，然后，用栅格法建立了本文AUV路径规划研究的航行环境模型，接下来分步从染色体编码、种群初始化、适应度函数、遗传算子的选择四个部分，对遗传算法的进行详细的改进。

最后，通过Matlab仿真验证，本文所提出的改进遗传算法能够使得AUV路径规划获得较优的路径规划结果。遗传模拟退火算法优于基本遗传算法，不仅可以解决其在计算过程中会地道道局部最优结果的问题，同时还可以提高收敛速度，从而加快AUV路径搜索的速度，使得AUV更能够满足现今社会的需求。

关键词：智能型水下机器人；路径规划；改进遗传算法；模拟退火

Abstract

The ocean contains abundant resources, which can supple the scarce energy on land today, so people begin to develop the ocean in depth. With the increasing demand for intelligent underwater robots in the process of exploring the ocean, the level of intelligence of AUV should also be constantly improved. Therefore, the control system of AUV movement and path planning of AUV are particularly important.

This paper mainly studies the path planning of AUV. In the second chapter of this paper, the AUV path planning problem is briefly described and analyzed, and then several common AUV path planning algorithms are introduced, and the corresponding research results at home and abroad are introduced. Then to study in this paper, including its solving steps in detail in this paper, the genetic algorithm, as will appear in the process of genetic algorithm in solving the situation of the local optimum, no longer continue to look for the optimal solution, so in this article puts forward the idea of simulated annealing algorithm is introduced into the genetic algorithm, the improved genetic algorithm for genetic simulated annealing algorithm. In the third chapter firstly introduces five kinds of environment modeling methods, respectively, but the view method, Voronoi method, grid method, unit tree method and topological method, and then by using the grid method established in this paper, the AUV navigation environment model of path planning study, the next step by step from the chromosome coding, initialization of population, fitness function and genetic operators choose four parts, to the improvement of genetic algorithm in detail.

Finally, through Matlab simulation verification, the improved genetic algorithm proposed in this paper can enable AUV path planning to obtain better path planning results. The genetic simulated annealing algorithm is better than the basic genetic algorithm, which can not only solve the problem that the algorithm will fall into the local optimum in the calculation process, but also improve the convergence speed, so as to speed up the AUV path search, so that the AUV can better meet the needs of today's society.

Key Words：AUV；Path Planning；Improved Genetic Algorithm；Simulated Annealing

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

1.2.1国外研究现状 2

1.2.2国内研究现状 3

1.3 本文研究内容 4

1.4 本文组织结构 4

第2章 路径规划算法的理论与研究 6

2.1 国内外AUV路径规划算法的研究 6

2.2 本文相关算法研究 7

2.2.1遗传算法 7

2.2.2模拟退火算法 8

2.2.3遗传模拟退火算法 10

2.3 本章小结 10

第3章 AUV路径规划研究的模型建立与算法设计 11

3.1 环境建模 11

3.1.1环境建模方法研究 11

3.1.2环境模型的建立 13

3.2 算法设计 14

3.2.1染色体编码 14

3.2.2种群初始化 15

3.2.3建立适应度函数 16

3.2.4遗传操作 18

3.2.5改进遗传算法的流程图 20

3.3 本章小结 20

第4章 仿真验证 21

4.1 不同参数的影响 21

4.2 基本与改进遗传算法的比较分析 23

4.3 经济环境分析 27

4.4 本章小结 27

第5章 结论 28

参考文献 29

致 谢 31

第1章 绪论

1.1 研究背景及意义

海洋本身是一个巨大的宝库，其中蕴藏着大量可再生性能源与稀缺资源，包括海洋中的生物，海水中的元素，波浪或潮汐所产生的能量、以及相当丰富的矿产资源。海洋中重要金属的含量是陆地矿产资源的数十倍以上，石油资源约为地球总石油量的百分之三十。这些能够补充现今匮乏的陆地资源，促进人类社会的进步与发展，因此海洋被称为21世纪人类可持续发展战略的替代者。