Real time automated-guided vehicle path control

An automated material handling system is an important component of most automated manufacturing systems. Automated Guided Vehicle Systems (AGVS) are commonly employed for this function. Since the AGV system plays a key role in factory automation, it has recently attracted tremendous attention. In this study, polynomial-time collision-free AGV path control algorithms and methods were developed, particularly for mid and large-sized problems. As the number of AGVs and complexity of AGV paths are greater, more intelligent and efficient AGV path control technologies are required. The algorithms have commitments to find the shortest time AGV path from one station to another while avoiding AGV collisions (blockings). They can detect both potential collisions and deadlocks in advance and avoid them. They are capable of responding in real time to the dynamic manufacturing environment that reflects any changes in manufacturing. A concept of symbolic time windows was developed here. Symbolic time windows take into account the current traffic situations in an AGV path network. Hence, the algorithms reason with symbolic time windows instead of the physical aspects of an AGV path network. All collision types such as head-on, catch-up, and node collisions are handled in this study. Therefore single-track bi-directional AGV path finding problems which are known to be more difficult can be solved. In this study, more than one vehicle is allowed to travel simultaneously in the same direction in an arc and multiple-speed types of AGVs are considered. Strategies for avoiding collisions and deadlocks are introduced in this study. The algorithms can provide both shortest-time (optimal) path and feasible paths to be used for the case where the shortest-time path cannot be obtained in real time. The algorithms and methods presented here are expected to replace the zone control method which is the state-of-the-art. They can be used for forward sensing which is one of the traffic control methods that use camera-based machine vision systems, ultrasonic sensors, lasers, and so on. They can be used for both guided-path AGVs and free-ranging (autonomous) AGVs. Finally, a computer simulation software prototype was developed to investigate the algorithms and strategies introduced in this research.