嵌入房间类别和边界约束的目标驱动导航算法

doi:10.19678/j.issn.1000-3428.0069313

摘要/Abstract

摘要：

在室内环境中, 不同房间类别的相同物体具有完全不同的用途, 设计附加房间类别约束的目标驱动导航任务, 在机器人导航、智能家居等领域具有重要应用。为提高房间类别约束目标导航任务的成功率, 设计一种结合映射模块、搜索策略、运动控制策略和房间分类模块的模块化导航算法。输入导航任务后, 映射模块结合RGB-D相机数据和惯导获得的姿态信息在线构建语义地图, 用于记忆已探索过的环境。在基于近端策略优化算法(PPO)框架实现搜索策略时, 为尽快找到地图上最可能存在目标物的坐标, 提出边界点簇的概念, 将其中心坐标作为中继点, 根据每个簇包含的边界点数量评定其中心点探索价值并排序, 用于约束全局目标点, 同时在搜索策略奖励函数中引入边界点约束, 以提高目标点落入已探索区域时的搜索效率。在基于运动控制策略控制机器人向着全局目标点移动的过程中, 针对机器人无法识别房间类别的问题, 采用YOLOv8_cls训练得到房间分类模块辅助进行决策, 从而更好地完成导航任务。分别在仿真环境和现实环境中验证导航任务的可完成性以及算法的有效性。实验结果表明, 所提出的算法相比于同样使用深度强化学习(DRL)作为搜索策略的SemExp (Semantic Exploration)算法, 在未附加以及附加房间类别约束的两类导航任务上, 能够更快地完成地图探索并且导航成功率分别提高2.0%和4.0%, 该算法能够更好地理解环境的语义信息, 完成未知环境中的目标物搜索等导航任务。

关键词: 机器人室内导航, 目标驱动, 房间类别约束, 搜索策略, 边界点约束

Abstract:

In indoor environments, the same object may have completely different uses depending on the room category. Thus, designing target-driven navigation tasks with room category constraints has important applications in robot navigation, smart home, and other fields. To improve the success rate of room category constrained target navigation task, a modular navigation algorithm is designed, combining search and motion control strategies with mapping and room classification modules. Given a navigation task as input, the mapping module combines RGB-D camera data and pose information, to construct an online semantic map that remembers environments that have been explored. The concept of boundary point cluster is proposed to quickly locate the most likely coordinates of the target object on the map as soon as possible when implementing the search strategy based on the proximal policy optimization algorithm framework. The central coordinates of these clusters are used as relay points. According to the number of boundary points contained in each cluster, the exploration value of the central point is evaluated and sorted and used to constrain the global target points. Furthermore, the concept of boundary points is introduced into the reward function of the search policy, to improve the search efficiency when the target points fall within the explored area. In response to the issue of the robot's inability to recognize room categories, YOLOv8_cls is trained to develop a room classification module based on the motion control strategy, to guide the robot towards the global target point to assist in decision-making, thereby better fulfilling navigation requirements. The feasibility of the navigation task and the effectiveness of the algorithm were verified in both simulated and real environments. Experimental results demonstrate that compared to the Semantic Exploration (SemExp) algorithm which employs Deep Reinforcement Learning (DRL) for search strategies, The proposed algorithm achieves faster map exploration and increased navigation success rates for two types of navigation tasks, with and without room category constraints by 2.0% and 4.0%, respectively. It demonstrates a better understanding of semantic information in the environment, enabling the completion of navigation tasks such as target object search in unknown environments.

Key words: robot indoor navigation, target-driven, room category constraint, search strategy, boundary point constraint

罗锦源, 谷雨. 嵌入房间类别和边界约束的目标驱动导航算法[J]. 计算机工程, 2025, 51(4): 85-96.

LUO Jinyuan, GU Yu. Target-driven Navigation Algorithm Embedded with Room Category and Boundary Constraints[J]. Computer Engineering, 2025, 51(4): 85-96.

https://www.ecice06.com/CN/Y2025/V51/I4/85

图/表 24

图1 目标驱动的导航任务描述

Fig.1 Description of target-driven navigation task

图2 导航算法流程

Fig.2 Process of navigation algorithm

图3 语义映射

Fig.3 Semantic mapping

图4 搜索策略网络结构

Fig.4 Network structure of the search strategy

图5 地图数据分析

Fig.5 Map data analysis

图6 全局目标选择

Fig.6 Selection of global target

图7 多余奖励示意图

Fig.7 Schematic diagram of surplus rewards

图8 运动控制策略

Fig.8 Motor control strategy

图9 麦克纳姆轮移动机器人

Fig.9 Mecanum wheel mobile robot

图10 房间分类结果

Fig.10 Result of room classification

图11 引入边界点约束的导航轨迹

Fig.11 Navigation trajectory with the introduction of boundary point constraints

图12 无边界点约束的导航轨迹

Fig.12 Navigation trajectory without boundary point constraints

图13 实验环境

Fig.13 Experimental environment

图14 深度图像的对比

Fig.14 Comparison of depth images

图15 实际部署的导航算法流程

Fig.15 Process of actual deployed navigation algorithm

图16 在线导航地图创建示意图

Fig.16 Schematic diagram of online navigation map creation

图17 一次导航的完整路径

Fig.17 Complete path of a navigation

图18 导航失败案例分析

Fig.18 Case analysis of navigation failure

参考文献 25

1	YE X, LIN Z, LI H X, et al. Active object perceiver: recognition-guided policy learning for object searching on mobile robots[C]//Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Washington D. C., USA: IEEE Press, 2018: 6857-6863.
2	ZHU Y K, MOTTAGHI R, KOLVE E, et al. Target-driven visual navigation in indoor scenes using deep reinforcement learning[C]//Proceedings of the IEEE International Conference on Robotics and Automation (ICRA). Washington D. C., USA: IEEE Press, 2017: 3357-3364.
3	王少桐, 况立群, 韩慧妍, 等. 基于优势后见经验回放的强化学习导航方法. 计算机工程, 2024, 50(1): 313- 319. doi: 10.19678/j.issn.1000-3428.0066193
	WANG S T, KUANG L Q, HAN H Y, et al. Reinforcement learning navigation method based on advantage hindsight experience replay. Computer Engineering, 2024, 50(1): 313- 319. doi: 10.19678/j.issn.1000-3428.0066193
4	YE J, BATRA D, DAS A, et al. Auxiliary tasks and exploration enable objectgoal navigation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2021: 16117-16126.
5	MAKSYMETS O, CARTILLIER V, GOKASLAN A, et al. THDA: treasure hunt data augmentation for semantic navigation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2021: 15374-15383.
6	户高铭, 蔡克卫, 王芳, 等. 基于深度强化学习的无地图移动机器人导航. 控制与决策, 2024, 39(3): 985- 993.
	HU G M, CAI K W, WANG F, et al. Mapless navigation based on deep reinforcement learning for mobile robots. Control and Decision, 2024, 39(3): 985- 993.
7	王少桐. 强化学习在机器人室内导航中的应用研究[D]. 太原: 中北大学, 2023.
	WANG S T. Research on the application of reinforcement learning in robot indoor navigation[D]. Taiyuan: North University of China, 2023. (in Chinese)
8	LIANG Y Q, CHEN B Y, SONG S R. SSCNav: confidence-aware semantic scene completion for visual semantic navigation[C]//Proceedings of the IEEE International Conference on Robotics and Automation (ICRA). Washington D. C., USA: IEEE Press, 2021: 13194-13200.
9	CHAPLOT D S, GANDHI D, GUPTA A, et al. Object goal navigation using goal-oriented semantic exploration[C]//Proceedings of the 34th International Conference on Neural Information Processing Systems (NeurIPS). New York, USA: ACM Press, 2020: 4247-4258.
10	RAMAKRISHNAN S K, CHAPLOT D S, AL-HALAH Z, et al. PONI: potential functions for objectgoal navigation with interaction-free learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D. C., USA: IEEE Press, 2022: 18890-18900.
11	YU B, KASAEI H, CAO M. L3MVN: leveraging large language models for visual target navigation[EB/OL]. [2023-04-11]. https://arxiv.org/abs/2304.05501.pdf.
12	XIA F, ZAMIR A R, HE Z Y, et al. Gibson Env: real-world perception for embodied agents[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D. C., USA: IEEE Press, 2018: 9068-9079.
13	HE K M, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2017: 2961-2969.
14	SCHULMAN J, WOLSKI F, DHARIWAL P, et al. Proximal policy optimization algorithms[EB/OL]. [2023-04-11]. https://arxiv.org/abs/1707.06347.pdf.
15	SUZUKI S, BE K. Topological structural analysis of digitized binary images by border following. Computer Vision, Graphics, and Image Processing, 1985, 30(1): 32- 46. doi: 10.1016/0734-189X(85)90016-7
16	VALERO-GOMEZ A, GOMEZ J V, GARRIDO S, et al. The path to efficiency: fast marching method for safer, more efficient mobile robot trajectories. IEEE Robotics&Automation Magazine, 2013, 20(4): 111- 120.
17	ARMENI I, HE Z Y, ZAMIR A, et al. 3D scene graph: a structure for unified semantics, 3D space, and camera[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2019: 5664-5673.
18	SAVVA M, KADIAN A, MAKSYMETS O, et al. Habitat: a platform for embodied AI research[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2019: 9339-9347.
19	LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: common objects in context[C]//Proceedings of the 13th European Conference on Computer Vision (ECCV). Berlin, Germany: Springer International Publishing, 2014: 740-755.
20	ANDERSON P, CHANG A, CHAPLOT D S, et al. On evaluation of embodied navigation agents[EB/OL]. [2023-04-11]. https://arxiv.org/abs/1807.06757.pdf.
21	ZHOU B L, LAPEDRIZA A, KHOSLA A, et al. Places: a 10 million image database for scene recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(6): 1452- 1464. doi: 10.1109/TPAMI.2017.2723009
22	HOWARD A, SANDLER M, CHEN B, et al. Searching for MobileNetV3[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Washington D. C., USA: IEEE Press, 2019: 1314-1324.
23	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Washington D. C., USA: IEEE Press, 2016: 770-778.
24	YAMAUCHI B. A frontier-based approach for autonomous exploration[C]//Proceedings of the IEEE International Symposium on Computational Intelligence in Robotics and Automation (CIRA). Washington D. C., USA: IEEE Press, 1997: 146-151.
25	何俊学, 李战明. 基于视觉的同时定位与地图构建方法综述. 计算机应用研究, 2010, 27(8): 2839- 2844.
	HE J X, LI Z M. Survey of vision-based approach to simultaneous localization and mapping. Application Research of Computers, 2010, 27(8): 2839- 2844.

[1]	向海昀, 李鸿鑫, 符晓, 苏小平. 基于多策略的改进蜜獾算法及其应用[J]. 计算机工程, 2023, 49(12): 78-87.
[2]	尚迪雅, 孙华, 洪振厚, 曾庆亮. 基于无梯度进化的神经架构搜索算法研究综述[J]. 计算机工程, 2020, 46(9): 16-26.
[3]	王斌, 房新秀, 魏天佑. 基于差异节点集的加权频繁项集挖掘算法[J]. 计算机工程, 2020, 46(5): 150-156.
[4]	高峰,刘震,高辉. 结合有监督广度优先搜索策略的通用垂直爬虫方法[J]. 计算机工程, 2018, 44(11): 289-299.
[5]	张琪, 张金林, 芮挺, 方虎生. 基于改进SIFT算法的粒子滤波目标跟踪[J]. 计算机工程, 2012, 38(10): 14-17.
[6]	王艺, 陈耀武. 一种视频转码快速运动估计算法[J]. 计算机工程, 2011, 37(12): 251-254.
[7]	张利群. 五道棋计算机博弈程序的设计与实现[J]. 计算机工程, 2010, 36(10): 221-222.
[8]	杨红;杨德礼;林正奎. 基于概念的软件体系结构需求目标描述与识别[J]. 计算机工程, 2007, 33(23): 60-62.

选择文件类型/文献管理软件名称

选择包含的内容