智慧道路施工机器人路径规划与控制方法研究

 2022-05-12 08:05

论文总字数:24414字

摘 要

根据十九大首次提出的发展“交通强国”的理念,全国高等级公路建设将迎来新一轮的建设高峰期。为了满足我国公路建设的巨大需求,缓解我国偏远地区公路建设的用工荒情况,减少人为造成的偷工减料和施工事故,针对人工智能化施工和工程标准化作业的研究是十分有必要的。本文提出的智慧道路施工机器人即是一种完全取代人工控制压路机作业的智能机器人。本文针对道路施工中的无人驾驶压路机的路径规划和控制方法进行了研究。论文主要内容如下:

  1. 研究了施工区域边界识别的方法,为准确的路径规划提供了理论基础。通过试验常见的K-means聚类方法,发现该方法无法很好地解决按多边形各边进行聚类的问题,导致后续的路径规划步骤无法进行。因此,本文根据聚类思想,提出了一种基于滑动窗口的边界聚类方法,并进行多次Matlab仿真实验验证。该方法将聚类的目标由点的坐标替换成滑动窗口内点集的二次拟合程度,将二次拟合效果差的连续窗口归为一类,通过对这些窗口的处理可以得到可疑拐点的范围,进而完成对边界识别。同时,在对边界的拟合环节,分情况讨论了矩形和扇形施工区域的拟合方法,发现扇形施工区域曲边的圆弧拟合比二次拟合效果好,提高了边界拟合的准确性。
  2. 在路径规划部分,根据我国压路机施工的规范,应考虑到相邻两条碾压路径的重叠标准,设置合适的碾压间距。同时,针对道路两侧路肩的充分压实问题,本文提出先沿一条侧边向另一侧碾压,直到即将超出施工范围,再沿另一条侧边反向碾压,此方法可以保证施工区域的全覆盖碾压,保证了施工质量。
  3. 针对施工机器人的轨迹跟踪问题,本文采取间接式航迹控制结构,即采用基于视线导向法(Line Of Sight Guidance,LOS)的导向算法确定下一时刻期望的航向角;采用基于PID算法的航向控制器跟踪此航向。最后,对此轨迹跟踪控制方法进行了Matlab仿真验证和实车实验以证实其可行性。

关键词:施工机器人,边界识别,全覆盖路径规划,视线导向法,PID控制算法

ABSTRACT

According to the concept of developing a “transportation power” proposed on the 19th National Congress, the construction of highways nationwide will enter a stage of rapid development. In order to meet the huge demand of highway construction in China, to alleviate the shortage of labor in highway construction in remote areas in China, and to reduce the man-made cuts and construction accidents caused by man-made, it is very necessary to study the artificial intelligent construction and engineering standardization operations. This paper studies the path planning and control methods of unmanned road rollers in road construction. The main contents of the paper are as follows:

1. The method of boundary identification of construction area is studied, which provides a theoretical basis for accurate path planning. By testing the common K-means clustering method, it is found that this method can not solve the problem of clustering of polygon sides. Therefore, based on the clustering idea, this paper proposes a boundary clustering method based on sliding window, and carries out several simulation experiments. The method replaces the target of the cluster with the coordinates of the point to the degree of quadratic fitting of the points in the sliding window, and classifies the continuous windows with poor quadratic fitting into one class. The suspicious inflection point can be obtained by processing the windows. The scope of the completion of the identification of the boundary. At the same time, in the fitting process of the boundary, the fitting method of the rectangular and fan-shaped construction areas is discussed. It is found that the arc-fitting of the sector-shaped construction area is better than the quadratic fitting, which improves the accuracy of the boundary fitting.

2. In the path planning part, according to the specification of roller construction in China, the wheel width overlap standard of two adjacent rolling paths should be considered, and the appropriate rolling distance should be set. At the same time, in view of the full compaction problem of the shoulders on both sides of the road, this paper proposes to first roll along one side to the other side until it is about to exceed the construction range, and then crush along the other side. This method can guarantee the construction area. The full coverage of the rolling, to ensure the quality of construction.

3. For the trajectory tracking problem of construction robots, this paper adopts the indirect trajectory control structure, which uses steering algorithm based on the Line Of Sight Guidance (LOS) to determine the desired heading angle at the next moment; using PID algorithm to control

the roller to track this heading angle. Finally, the trajectory tracking control method is verified by simulation and experiments to confirm its feasibility.

KEY WORDS: construction robot, boundary recognition, full coverage path planning, line of sight guidance, PID control

目 录

摘要 3

ABSTRACT 4

第一章 绪论 8

1.1研究背景及意义 8

1.2智慧道路施工机器人研究现状 9

1.2.1国外研究现状 9

1.2.2国内研究现状 9

1.2.3研究现状小结 11

1.3论文内容安排 11

1.4本章小结 12

第二章 智慧道路施工机器人系统构成 13

2.1引言 13

2.2智慧道路施工机器人的系统框架 13

2.3智慧道路施工机器人控制系统硬件结构 14

2.4施工机器人路径规划与控制方法的技术路线 15

2.5本章小结 16

第三章 施工机器人路径规划方案设计 17

3.1引言 17

3.2基于滑动窗口的识别施工边界方案设计 17

3.2.1采集数据环节设计 17

3.2.2边界聚类环节设计 17

3.2.3解析拐点环节设计 19

3.2.4边界拟合环节设计 20

3.3路径规划环节设计 22

3.3.1路径规划初始化环节 23

3.3.2重复实压环节 23

3.4本章小结 24

第四章 施工机器人控制算法设计 25

4.1引言 25

4.2施工机器人运动学模型 25

4.3施工机器人控制系统结构 26

4.4导向算法 27

4.5航向控制器 28

4.5.1 PID控制算法简介 28

4.6本章小结 29

第五章 施工机器人路径规划与控制实验 30

5.1引言 30

5.2路径规划算法仿真 30

5.2.1直线路面路径规划 30

5.2.2曲线路面路径规划 31

5.3施工机器人轨迹跟踪实验 33

5.3.1仿真试验分析 33

5.3.2实车实验 36

第六章 总结与展望 38

6.1总结 38

6.2展望 38

参考文献 40

致 谢 42

第一章 绪论

1.1研究背景及意义

在十九大报告中,习总书记首次提出要加强“交通强国”建设,并表示要“加强水利、铁路、公路、水运、航空、管道、电网、信息、物流等基础设施网络建设”。2018年起,交通强国战略正式实施,据不完全统计,有22个省市公布了交通投资计划,累计额度超过了两万亿元。在巨大的公路建设需求下,部分偏远山区的用工荒情况愈发严重,与此同时,人为造成的偷工减料和施工事故也愈演愈烈,因此实现人工智能化施工,建立智能施工标准对实现“交通强国”和“智能制造2025”的国家战略目标具有重要的推动作用。

交通基建行业对国民经济发展和社会民生起着重要的基础性、支撑性和服务性的作用。 “一带一路”国家战略,前期核心任务是以基础建设为主。中国的公路建设经过20年的高速发展,积累了大量的经验,在新时代、人工智能高速发展的今天,如何实现道路高精准度、高安全性、高效率施工,将是我国施工单位今后相当长一段时间内发展的主旋律,更是我国打造先进交通建设“国家名片”,全面支撑“交通强国”、“中国制造2025”、“走出去”、“一带一路”、“区域经济一体化”战略的全局性重要基础保障。

剩余内容已隐藏,请支付后下载全文,论文总字数:24414字

您需要先支付 80元 才能查看全部内容!立即支付

该课题毕业论文、开题报告、外文翻译、程序设计、图纸设计等资料可联系客服协助查找;