论文阅读（一） Indoor Coverage Path Planning: Survey, Implementation, Analysis

Indoor Coverage Path Planning: Survey, Implementation, Analysis

概况：讲的是覆盖路径规划（CPP）算法，本文中介绍了六种算法，并进行了深入的比较，在550多个室内地图中做了实验。

 

一、介绍

本文是基于6\8论文的补充。

本文提供了广泛的覆盖路径规划方法，包括一种精确的单元分解方法、基于三单元网格的规划方法、一种轮廓线驱动方法和一种艺术画廊问题的变体。（不知道准不准确，英文是one exact cell decomposition method, three cell grid-based planners, one contour line-driven approach and one variant of the art gallery problem）

区分了在线算法（基于传感器的覆盖算法）和离线算法，本文的工作集中在离线算法上，这些算法可以利用环境的现成的地图。

 

二、相关的工作

总结讨论了6\8论文以及其他比较常见的算法：

　　经典的网格划分方法（论文10-12）：依赖于多边形形状和障碍物。解决这一问题的方法：Morse-based网格划分方法（论文13）。

　　Landmark-based网格划分方法（论文14）：适用于多种拓扑结构。

　　Boustrophedon网格划分方法（论文11\15）

　　旋转扫描单元格的Boustrophedon网格划分方法（论文16）　　

　　Wavefront 算法（论文17）：定义了起点和目标点，接近目标前将这拨钱

　　Spanning Tree算法（论文18）：将自由空间分割为巨型的单元，在每个巨型单元下有四个小单元，可以利用搜索树来覆盖这些单元。

　　Neural Network-based算法（论文19）：受生物学启发，将网格视作神经元，通过相邻网格的激活状态访问网格。

　　全覆盖问题被视作Art Gallery Problem（论文20）或者Watchman Tour Problem 。

　　在组合整数线性规划方法的基础上，利用重加权凸松弛扩大可处理规模的一种新方法（论文9）

从中挑选6种全覆盖路径算法进行评价。

 

三、方法

1、预处理

（1）地图分割。看作Traveling Salesman Problem，这篇论文的作者此前写过一篇相关论文7 。

（2）房间方向规范化。墙壁对准的启发式算法（论文6\12\16）、Canny边缘检测算法、CV的Hough直线检测算法（论文22）

（3）覆盖区域： skeleton or Voronoi graph 论文23\24。网格最多为l=√2·rq 。

2、  Boustrophedon Coverage Path Planning 论文11\15 。水平扫描，到达临界点时执行合并或者分割。

3、Grid-based Traveling Salesman Coverage Path Planning 。基于旅行商TSP问题。与Concorde solver完全一致（论文25），或与遗传算法或简单而快速的最近邻类似。

4、Neural Network-based Coverage Path Planning 。基于神经网络的全覆盖算法。结合生物学的神经元连接8个邻居，排斥墙壁和障碍物。论文19

5、Grid-based Local Energy Minimization。局部能量最小驱动。 论文5

6、Contour Line-based Coverage Path Planning。等高线法。论文7\26

7、Convex Sensor Placement Coverage Path Planning。凸传感覆盖算法。论文9

 

四、评估

1、覆盖率 96-99%

2、传感器视野

3、实测结果

 

 

【延伸论文】

[1] F. Yasutomi, M. Yamada, and K. Tsukamoto, “Cleaning robot control,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), Apr. 1988, pp. 1839–1841.
[2] C. Hofner and G. Schmidt, “Path planning and guidance techniques for an autonomous mobile cleaning robot,” Robotics and Autonomous Systems, vol. 14, no. 23, pp. 199 – 212, 1995.
[3] M. Bosse, N. Nourani-Vatani, and J. Roberts, “Coverage Algorithms for an Under-actuated Car-Like Vehicle in an Uncertain Environment,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), Apr. 2007, pp. 698–703.
[4] M. Ollis and A. Stentz, “Vision-based perception for an automated harvester,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), vol. 3, Sept. 1997, pp. 1838– 1844.
[5] R. Bormann, J. Hampp, and M. Hagele, “New Brooms Sweep Clean ¨ - An Autonomous Robotic Cleaning Assistant for Professional Office Cleaning,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), May 2015.
[6] E. Galceran and M. Carreras, “A survey on coverage path planning for robotics,” Robotics and Autonomous Systems, vol. 61, no. 12, pp. 1258–1276, 2013. [Online]. Available: http://www.sciencedirect.com/ science/article/pii/S092188901300167X
[7] R. Bormann, F. Jordan, W. Li, J. Hampp, and M. Hagele, “Room ¨ Segmentation: Survey, Implementation, and Analysis,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2016.

[8] H. Choset, “Coverage for robotics - A survey of recent results,” Annals of mathematics and artificial intelligence, vol. 31, no. 1, pp. 113–126, 2001. [Online]. Available: http://www.springerlink.com/index/q346pp34036143pg.pdf
[9] M. A. Arain, M. Cirillo, V. H. Bennetts, E. Schaffernicht,M. Trincavelli, and A. J. Lilienthal, “Efficient measurement planning for remote gas sensing with mobile robots,” in Proceedings of the IEEE International Conference on Robotics and Automation
(ICRA). IEEE, 2015, pp. 3428–3434. [Online]. Available: http://ieeexplore.ieee.org/abstract/document/7139673/
[10] J.-C. Latombe, “Exact Cell Decomposition,” in Robot Motion Planning, ser. The Springer International Series in Engineering and Computer Science. Springer US, 1991, no. 124, pp. 200–247, dOI: 10.1007/978-1-4615-4022-9 5. [Online]. Available:
http://link.springer.com/chapter/10.1007/978-1-4615-4022-9 5
[11] H. Choset and P. Pignon, “Coverage path planning: The boustrophedon cellular decomposition,” in Proceedings of the International Conference on Field and Service Robotics. Springer, 1998, pp. 203–209. [Online]. Available: http://link.springer.com/chapter/10.1007/978-1-4471-1273-0 32
[12] T. Oksanen and A. Visala, “Coverage path planning algorithms for agricultural field machines,” Journal of Field Robotics, vol. 26, no. 8, pp. 651–668, Aug. 2009. [Online]. Available: http://onlinelibrary.wiley.com/doi/10.1002/rob.20300/abstract
[13] E. U. Acar, H. Choset, A. A. Rizzi, P. N. Atkar, and D. Hull,“Morse Decompositions for Coverage Tasks,” The International Journal of Robotics Research, vol. 21, no. 4, pp. 331–344, Apr. 2002. [Online]. Available: http://journals.sagepub.com/doi/abs/
10.1177/027836402320556359
[14] S. C. Wong and B. A. MacDonald, “A topological coverage algorithm for mobile robots,” in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), vol. 2, Oct. 2003, pp. 1685–1690.
[15] H. Choset, E. Acar, A. A. Rizzi, and J. Luntz, “Exact cellular decompositions in terms of critical points of morse functions,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), vol. 3. IEEE, 2000, pp. 2270–2277. [Online]. Available: http://ieeexplore.ieee.org/abstract/document/846365/
[16] W. H. Huang, “Optimal line-sweep-based decompositions for coverage algorithms,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), vol. 1, 2001, pp. 27–32. [17] A. Zelinsky, R. A. Jarvis, J. C. Byrne, and S. Yuta, “Planning Paths of Complete Coverage of an Unstructured Environment by a Mobile Robot,” in In Proceedings of International Conference on Advanced Robotics, 1993, pp. 533–538. [18] Y. Gabriely and E. Rimon, “Spiral-STC: an on-line coverage algorithm of grid environments by a mobile robot,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), vol. 1,2002, pp. 954–960.
[19] S. X. Yang and C. Luo, “A neural network approach to complete coverage path planning,” IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), vol. 34, no. 1, pp. 718–724, Feb. 2004.
[20] T. Shermer, “Recent results in art galleries,” Proceedings of the IEEE, vol. 80, no. 9, pp. 1384–1399, Sep 1992. [21] “Room segmentation dataset and software,” http://wiki.ros.org/ipa room segmentation.
[22] G. Bradski and A. Kaehler, Learning OpenCV: Computer vision with the OpenCV library. ” O’Reilly Media, Inc.”, 2008.
[23] S. Fortune, “A sweepline algorithm for voronoi diagrams,” Algorithmica, vol. 2, no. 1-4, pp. 153–174, 1987.
[24] H. Choset, “Incremental construction of the generalized voronoi diagram, the generalized voronoi graph, and the hierarchical generalized voronoi graph,” in Proceedings of the First CGC Workshop on Computational Geometry, 1997.
[25] D. Applegate, R. Bixby, V. Chvatal, and W. Cook, “Concorde TSP solver,” http://www.math.uwaterloo.ca/tsp/concorde.html, 2006. [26] S. Thrun, “Learning metric-topological maps for indoor mobile robot navigation,” Artificial Intelligence, vol. 99, no. 1, pp. 21–71, Feb. 1998. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0004370297000787
[27] “Documentation on ipa room exploration software,” http://wiki.ros. org/ipa room exploration.