Automated Guided Vehicles (AGVs) are driverless vehicles that are becoming a popular means for material handling. Existing methods for routing AGVs are generally based on centralized control. The central controller is potentially a bottleneck of communication and the single point of weakness of the system. Moreover, the computational cost required to preplan all AGV routes may be too high as to affect the scalability of the solution. Our project aims to design and develop decentralized routing strategies for an AGV system deployed within a container port. Two routing algorithms were developed: Row-First and Semi-Random. The goal is to minimize deadlocks and congestion and maximize throughput. The performance of the algorithms were monitored and compared with each other using a graphics based computer simulator called SIM-PORT.
[1]
Arie Shoshani,et al.
System Deadlocks
,
1971,
CSUR.
[2]
Yavuz A. Bozer,et al.
Tandem Configurations for Automated Guided Vehicle Systems and the Analysis of Single Vehicle Loops
,
1991
.
[3]
Stephen C. Daniels.
Real Time Conflict Resolution in Automated Guided Vehicle Scheduling
,
1988
.
[4]
Hsu Weng Jing.
Two Arms Are Faster Than One
,
1994
.
[5]
John J. Bartholdi,et al.
Decentralized Control of Automated Guided Vehicles on a Simple Loop
,
1989
.
[6]
Hsu Wen Jing,et al.
Route planning of automated guided vehicles
,
1994,
Proceedings of the Intelligent Vehicles '94 Symposium.