MODULE BASED STUDIES ON LARGE SCALE AGV SYSTEM IN AUTOMATED TRANSSHIPMENT CONTAINER TERMINAL

Abstract

In large-scale automated transshipment container terminals, the overall traffic network can be divided into different types of modules which are large enough to hold many AGVs. After decomposition of the modules, the AGV routing and scheduling problem can be divided into two sub-problems: the rough routes generation problem which determines the shortest time path for all AGVs and avoid possible congestions, and inner module control problem which determines the routes and schedules for all the AGVs inside the module to minimize overall travel time and avoid conflicts and deadlocks. These two problems are integrated together to achieve real time large scale AGV system control. The modules can also be simulated as servers with certain service rates. The server network is integrated with the yard system to find optimal yard allocation decisions to reduce the overall AGV traffic time, and therefore improve the overall efficiency of the whole container terminal.