Warehouse AGV Navigation Based on Multimodal Information Fusion

Bo Zhang; Yinlong Zhang; Wei Liang; Xin Wang; Yutuo Yang

doi:10.3788/AOS231613

Acta Optica Sinica, Volume. 44, Issue 9, 0915003(2024)

Warehouse AGV Navigation Based on Multimodal Information Fusion

Bo Zhang¹, Yinlong Zhang^2,3,4、*, Wei Liang^2,3,4, Xin Wang¹, and Yutuo Yang^2,3,4

Author Affiliations

¹School of Electrical and Control Engineering, Shenyang Jianzhu University, Shenyang 110168, Liaoning, China

²Key Laboratory of Networked Control System, Chinese Academy of Sciences, Shenyang 110169, Liaoning, China

³Institute of Robotics and Intelligent Manufacturing Innovation, Chinese Academy of Sciences, Shenyang 110169, Liaoning, China

⁴State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110169, Liaoning, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. Laser-inertial QR code fusion-based AGV navigation framework

Download full size

Fig. 2. Set of points formed by a corner point and its neighbors and set of points formed by a plane point and its neighbors. (a) Corner point and its neighbors; (b) plane point and its neighbors

Download full size

Fig. 3. Schematic diagram of hierarchical BA with QR code fusion

Download full size

Fig. 4. Factor graph structure diagram

Download full size

Fig. 5. Simulated warehouse workshop scene layout and experimental hardware platform. (a) Simulation of warehouse workshop scene; (b) experimental hardware platform

Download full size

Fig. 6. Experimental result of line trajectory and comparison of different SLAM methods. (a) Linear experiment effect; (b) comparison of line trajectories of different SLAM methods

Download full size

Fig. 7. Analysis of absolute trajectory error for long-term data of linear trajectories. (a) Absolute translation error of linear trajectories; (b) absolute rotation error of linear trajectories

Download full size

Fig. 8. Experimental results of rectangle trajectory and comparison of different SLAM methods. (a) Rectangular experiment effect; (b) comparison of rectangle trajectories of different SLAM methods

Download full size

Fig. 9. Analysis of absolute trajectory error for long-term data of rectangle trajectories. (a) Absolute translation error of rectangular trajectories; (b) absolute rotation error of rectangular trajectories

Download full size

Table 1. Hierarchical BA parameter settings
View table
Table 1. Hierarchical BA parameter settings
Parameter Description Setting
$ω$ Window size 12
$s$ Stride 6
$n$ Parallel thread 8

Table 2. Comparison of absolute trajectory error for line trajectories

View table

Table 2. Comparison of absolute trajectory error for line trajectories

Trajectory	Length /mm	Metric	LeGO-LOAM	BALM	LIO-SAM	LIC-Fusion2	Ours
Line	24000	$A_{t r a n s}$	332.85 mm	295.52 mm	75.56 mm	54.34 mm	14.56 mm
	24000	$A_{r o t}$	3.61°	2.35°	2.12°	1.77°	1.12°
	36000	$A_{t r a n s}$	343.26 mm	301.72 mm	77.78 mm	58.55 mm	15.95 mm
	36000	$A_{r o t}$	4.15°	2.92°	2.27°	1.85°	1.13°
	48000	$A_{t r a n s}$	346.65 mm	307.25 mm	86.23 mm	65.78 mm	17.27 mm
	48000	$A_{r o t}$	4.52°	3.17°	2.76°	2.16°	1.35°
	60000	$A_{t r a n s}$	351.35 mm	309.67 mm	90.74 mm	69.91 mm	18.38 mm
	60000	$A_{r o t}$	5.17°	3.53°	2.98°	2.49°	1.63°

Table 3. Comparison of absolute trajectory error for rectangle trajectories

View table

Table 3. Comparison of absolute trajectory error for rectangle trajectories

Trajectory	Length /mm	Metric	LeGO-LOAM	BALM	LIO-SAM	LIC-Fusion2	Ours
Rectangle	16800	$A_{t r a n s}$	226.36 mm	141.73 mm	48.77 mm	28.41 mm	10.96 mm
	16800	$A_{r o t}$	4.61°	5.67°	3.32°	2.37°	1.32°
	33600	$A_{t r a n s}$	239.22 mm	153.65 mm	51.34 mm	37.26 mm	12.47 mm
	33600	$A_{r o t}$	5.15°	6.76°	4.27°	2.91°	1.83°
	50400	$A_{t r a n s}$	245.89 mm	158.42 mm	55.62 mm	43.35 mm	11.52 mm
	50400	$A_{r o t}$	5.92°	7.46°	3.76°	3.38°	1.96°
	67200	$A_{t r a n s}$	252.94 mm	161.86 mm	61.15 mm	49.74 mm	13.27 mm
	67200	$A_{r o t}$	6.77°	8.34°	4.98°	3.76°	1.98°

Table 4. Comparison of time efficiency of threads

View table

Table 4. Comparison of time efficiency of threads

Thread	LeGO-LOAM	BALM	LIO-SAM	LIC-Fusion2	Ours
Date preprocessing	52.43	46.38	58.12	58.34	65.68
Laser odometry	40.12	31.22	28.36	27.17	25.64
Optimized mapping	653.34	627.35	234.86	462.15	140.14
Loop detection	368.32	—	382.95	351.49	332.31
System operation	1114.21	704.95	704.29	899.15	563.77

Tools

Get Citation

Copy Citation Text

Bo Zhang, Yinlong Zhang, Wei Liang, Xin Wang, Yutuo Yang. Warehouse AGV Navigation Based on Multimodal Information Fusion[J]. Acta Optica Sinica, 2024, 44(9): 0915003

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites