1-s2.0-s0305054810001760-main
DESCRIPTION
MainTRANSCRIPT
-
img
ro
on, G
3, U
po
at
rel
ster
pro
on
d a
ous
e t
from computational experiments.
1. Introduction
(FMSsindustatedto mtributithe syof prossue t
s fordingm isimeict-
et al. [7] provided a mathematical model for routing problem.
Contents lists available at ScienceDirect
journal homepage: www.e
Computers & Oper
Computers & Operations Research 38 (2011) 876888the case of failure.1 Tel.:+1 4122683642; fax:+1412 2687139.Lagrangian decomposition technique was used solve the problem.Ghasemzadeh et al. (2009) [8] presented a conict-free schedul-ing and routing in mesh topologies. It can generate the shortestpath for scheduling predicting conicts and select another path in
0305-0548/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cor.2010.08.012
Corresponding author. Tel./fax: +81668506351.E-mail addresses: [email protected] (T. Nishi),
[email protected] (I.E. Grossmann).AGVs for semiconductor fabrication. The interaction betweenproduction and transportation control is discussed by Mantel and
free conditions [4,5]. Singh and Tiwari [6] presented an intelligentagent framework to nd a conict-free shortest-time path. Nishi(1) dispatching, which is to assign tasks to vehicles; (2) routing,which is to select specic paths taken by vehicles; and(3) scheduling, which is to determine the arrival and departure times.
Unlike the classical vehicle routing problem (VRP) formulation,conict-free constraints should be considered for the routing of
performance taking into account for deadlock or conictAGVs. Kim and Tanchoco [3] studied the problem of nconict-free routes in a bidirectional network. The algorithbased on the shortest path methods through the concept of twindow graph. Petri net is used to analyze deadlock and conthe simultaneous optimization problems for production schedul-ing, dispatching and routing for vehicles in the static situation.The production scheduling problems asks an optimal productionsequence and starting time of operations for jobs at machines formulti-stages with respect to a specied technical precedencerelation. The vehicle management problems are classied into
Most of the literature treat one or two of the problems at the sametime. An extensive review has been addressed by Vis [2] foroperational control of AGVs. A widely used technique fordispatching is the simulation. The heuristic rules are used inon-line control systems. For routing and scheduling of AGVs,several techniques have been used to maximize the total systemFlexible manufacturing systemswarehousing systems, and servicetransportations are employing autom(AGVs) for the material handlingefciency of production and disoperation, it is requested to realizefor the simultaneous schedulingtransportation systems. The main i& 2010 Elsevier Ltd. All rights reserved.
), container terminals,ries including hospitalguided vehicle systemsaintain exibility andon. For the efcientnchronized operationsduction systems andreated in this paper is
Landerweerd [1]. In the owshop production systems, theproduction and transportation schedules are usually controlledby a pull type of policy in case of forklifts or conveyor systems.However, for FMSs environment with AGV systems, the optimalmachine schedules highly depend on the selection of dispatchingand routing because it is extremely difcult to predict thetransportation time when the conicts and interferences betweenvehicles cannot be neglected.
Production scheduling, dispatching, routing and schedulingdecisions for AGVs can be made simultaneously or separately.A bilevel decomposition algorithm for sand conict-free routing for automated
Tatsushi Nishi a,, Yuichiro Hiranaka a, Ignacio E. Ga Division of Mathematical Science for Social Systems, Department of Systems Innovati
Toyonaka City, Osaka 560-8531, Japanb Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA 1521
a r t i c l e i n f o
Available online 7 September 2010
Keywords:
Bilevel decomposition
Lagrangian relaxation
Scheduling
Routing
Automated guided vehicle
a b s t r a c t
We address a bilevel decom
routing problems for autom
tardiness of the set of jobs
levels: the upper level ma
subproblem. The master
obtained. Either the soluti
problem is constructed, an
generated to exclude previ
cuts are proposed to reducultaneous production schedulinguided vehicles
ssmann b,1
raduate School of Engineering Science, Osaka University, 1-3 Machikaneyama,
SA
sition algorithm for solving the simultaneous scheduling and conict-free
ed guided vehicles. The overall objective is to minimize the total weighted
ated to these tasks. A mixed integer formulation is decomposed into two
problem of task assignment and scheduling; and the lower level routing
blem is solved by using Lagrangian relaxation and a lower bound is
turns out to be feasible for the lower level or a feasible solution for the
n upper bound is obtained. If the convergence is not satised, cuts are
feasible solutions before solving the master problem again. Two types of
he duality gap. The effectiveness of the proposed method is investigated
lsevier.com/locate/caor
ations Research