Computer Engineering

Target-driven Navigation Algorithm Embedded with Room Category and Boundary Constraints

LUO Jinyuan, GU Yu

Computer Engineering, 2025, 51(4): 85-96. DOI: 10.19678/j.issn.1000-3428.0069313

Table 6 Ablation experiment result of boundary point constraint

Other figure/table from this article

Fig.1 Description of target-driven navigation task
Fig.2 Process of navigation algorithm
Fig.3 Semantic mapping
Fig.4 Network structure of the search strategy
Fig.5 Map data analysis
Fig.6 Selection of global target
Fig.7 Schematic diagram of surplus rewards
Fig.8 Motor control strategy
Table 1 Parameters setting for search strategy
Fig.9 Mecanum wheel mobile robot
Table 2 Performance comparison of different models on the room classification task
Fig.10 Result of room classification
Table 3 Parameters setting of reward function
Table 4 Comparative experiment result of traditional target-driven navigation task
Fig.11 Navigation trajectory with the introduction of boundary point constraints
Fig.12 Navigation trajectory without boundary point constraints
Table 5 Comparative experiment result of navigation tasks with room constraints
Fig.13 Experimental environment
Fig.14 Comparison of depth images
Fig.15 Process of actual deployed navigation algorithm
Fig.16 Schematic diagram of online navigation map creation
Fig.17 Complete path of a navigation
Fig.18 Case analysis of navigation failure