Petri‐net‐based deadlock detection and recovery for control of interacting equipment in automated container terminals

Abstract

In automated container terminals (ACTs), quay cranes (QCs), automated guided vehicles (AGVs), and automated yard cranes (AYCs) interact intensively to handle containers at seaside and yard side. This work investigates interacting equipment in ACTs that employ AGVs with the capability of lifting a container temporarily buffered at the transfer area in front of a storage block. By establishing and using the Petri net model of equipment interaction system (EIS) for container transportation, deadlock control problem is studied for handling operations in ACTs, where all QCs, AYCs, and transport vehicles are considered as system resources. Deadlock is characterized as maximal special circuits (MSCs) that are saturated at a system state. After a deadlock detection method through MSC enumeration in the system Petri net model is developed, a procedure is proposed to resolve deadlock situations where an AYC is involved in the corresponding circular waiting relations. By modifying the operation sequence of AYCs, the procedure can resolve the conflicts among concurrent processes so as to eliminate deadlocks. Examples are presented to illustrate the effectiveness of the deadlock handling method, which is also tested in a simulation study. Results show that the EIS can recover from detected deadlocks involving AYCs to ensure smooth transport operations of container jobs.

Emerging technology and management research in the container terminals: Trends and the COVID-19 pandemic impacts

“…Among all the research problems, deadlock- and collision-free scheduling and path planning for large scale vehicle fleet are the most challenging part in AGV operations, as the operators prefer to guarantee safety with less throughput rather than go to the ground to resolve physical issues. Although conflict-free constraint ( Hu et al, 2022 ) or deadlock detection method ( Wu et al, 2022 ) are considered, conflict and deadlock are still resolved at route level while vehicle maneuver and node-arc map are not captured. The studies on CTs pay more attention on planning level, such as truck dispatching/deployment ( Chargui et al, 2021 ) and truck appointment system for arrival scheduling (will be discussed in later section).…”

Intersection traffic deadlock formation and its probability: A petri net‐based modeling approach

“…In the transportation area, a PN model has also been successfully applied to signal control [17][18][19], copilot the system [20], automated guided vehicles [30,31], and others. As PN can provide the analytical formulation of deadlock state formation and is applied to traffic flow researches successfully [33][34][35], we employ the Petri Net (PN)-based modelling approach to investigate deadlock formation, including its formation probability and duration. In traditional PN design for intersection traffic flow or automated guided vehicles, the interactions rules is relative simple and cannot reproduce the deadlock successfully [17], and the environment is structured.…”

“…PN is a basic tool in discrete event system (DES) analysis and is applied to deadlock successfully [33, 35]. Mathematically, a PN is a 4‐tuple $$( {P,T,F,W} )$$, where P is the set of places and T is the set of transitions .…”