Gerrit K. Janssens

New heuristics for the fleet size and mix vehicle routing problem with time windows

In the Fleet Size and Mix Vehicle Routing Problem with Time Windows (FSMVRPTW) customers need to be serviced in their time windows at minimal costs by a heterogeneous fleet. In this paper new heuristics for the FSMVRPTW are developed. The performance of the heuristics is shown to be significantly higher than that of any previous heuristic approach and therefore likely to achieve better solutions to practical routing problems.

Data mining with genetic algorithms on binary trees

Abstract This paper focuses on the automatic interaction detection (AID)-technique, which belongs to the class of decision tree data mining techniques. The AID-technique explains the variance of a dependent variable through an exhaustive and repeated search of all possible relations between the (binary) predictor variables and the dependent variable. This search results in a tree in which non-terminal nodes represent the binary predictor variables, edges represent the possible values of these predictor variables and terminal nodes or leafs correspond to classes of subjects. Despite of being self-evident, the AID-technique has its weaknesses. To overcome these drawbacks a technique is developed that uses a genetic algorithm to find a set of diverse classification trees, all having a large explanatory power. From this set of trees, the data analyst is able to choose the tree that fulfils his requirements and does not suffer from the weaknesses of the AID-technique. The technique developed in this paper uses some specialised genetic operators that are devised to preserve the structure of the trees and to preserve high fitness from being destroyed. An implementation of the algorithm exists and is freely available. Some experiments were performed which show that the algorithm uses an intensification stage to find high-fitness trees. After that, a diversification stage recombines high-fitness building blocks to find a set of diverse solutions.

Techniques for Modeling Workflows and Their Support of Reuse

Several authors propose their own technique based on Petri Nets to model workflow processes. Most of them recognise the adaptability problem inherent to workflows, viz. the frequently and/or radically changing character due to changing business process rules, but suggest totally different solutions. Because the proposed techniques are fundamentally different, eleven of these techniques are briefly discussed and compared. Next, we survey approaches to reuse in the workflow field and we classify them in a framework derived from the information systems literature.

EDI—A strategic weapon in international trade

Abstract While electronic data interchange will in the long term be a necessity for industry there are problems to be solved before it is generally accepted. Close co-operation between European standard committees and the European community is recommended if European industry as a whole is not to fall behind in the development.

A Petri net model of a continuous flow transfer line with unreliable machines

Abstract A production line is studied with several machines in series. The machines are separated with finite buffers. They can break down and have to be repaired. Production is assumed to be continuous, i.e. no individual products can be identified. Analytical models for such a system are scarce. Analytical results for the throughput can be obtained only under very restrictive assumptions. The study of longer transfer lines requires either approximation or simulation. A general Petri net model is developed for an n-machine (n−1)-buffer system. Rules to dress up the Petri net transitions are given. The automatic generation of a discrete-event simulation model from the Petri net model is explained. Finally, some potential applications of the Petri net are illustrated.

Buffer allocation and required availability in a transfer line with unreliable machines

AbstractInthispaperwestudyasystemof n machinesinseries.Eachpairofsubsequentmachinesisseparatedbyanitebu!er.Themachinesareunreliableandcanbreakdown.Insuchaproductionsystem,theallocationofbu!erspacebetweensuccessivestagesinthesystemisofcapitalimportancesinceithasasignicantin#uenceontheavailabilityofthesystem.Weexaminehowtheallocationofbu!erscanbeexpressedasanon-linearoptimisationprobleminwhichthetotalcostof installing and using the bu!ers is minimised. Some of the assumptions of this approximation model are alsoexamined. 2001ElsevierScienceB.V.All rightsreserved. Keywords: Bu!er;Availability;Transferline;Unreliablemachines 1. IntroductionFlow-line production systems have receivedagreatdealofattentionintheliteratureaswellason the productionshop#oor.A #ow-lineproduc-tionsystemconsistsofalinearsequenceofseveralmachines.Sincethemachinescannotperformtheiroperationswhentheydonotreceiveinputorcan-not produce output, the entire production systemfailsifonemachinefails.Onesolutiontothisprob-lemistoinstallsecondary(i.e.standby)machinesinparallelwiththeprimarymachinesthatcomeintooperation when a machine fails. Another possiblesolutionistoinstallbu!erstoragebetweensuccess-ive stages of the production line, unlinking thesestages and allowing production to continue untilthe machines are repaired. An important decisioninthedesignofa#ow-lineproductionsystemistheamountofbu!erspacetoinstallbetweensuccessivestages. This decision is of great practical import-ance to such industries as bottling, packaging, as-semblingandautomotiveproduction.If thebu!erspaceinstalledbetweentwomachinesisinsu

Design, performance and wavelength assignment of a wavelength division multiaccess protocol for optical fibre ring networks

cienttherewillnotbeenoughtimetorepairthemachin-esbeforetheentire#ow-linefails.Toomuchbu!erspacewillincuracostthatishigherthannecessary.Mathematical models of #ow-line productionsystems can be divided into two categories: con-tinuous and discrete. In discrete #ow-line produc-tion systems individual parts are treated by thesuccessive stages of the line. Continuous #ow-linesystemsoperate on a #ow of the product throughthesystem.Thelatterapproachissuitableforpro-duction systems in which no individual items canbeidentied,suchasintheprocessindustry.Itcan

Opportunities of robust regression for variance reduction in discrete event simulation

This paper presents a multiaccess strategy for use over optical fibre ring networks employing two counter rotating rings, and using wavelength division multiplexing. Transmitters employ tunable lasers, while receivers employ one or more fixed wavelength filters. The paper introduces an algorithm for the optimal assignment of the receivers to the channels, such that the possibility of finding an unused channel leading to a certain receiver is maximized. The paper contains a performance study of the network protocol.

ASSESSING THE IMPACT OF EDI: A COST ACCOUNTING FRAMEWORK.

Variance reduction increases the efficiency of discrete event simulation. The method of using control variates searches for variables with known mean which are correlated with the response variable. The more correlated they are, the more the variance of the response variable can be reduced to obtain a confidence interval. Mostly only a few replications are made. This enables outliers to give a false view on the reduction possibilities. To cope with this statistical problem, we propose the use of a robust regression method in determining optimal weights for the control variables.

Outliers, Sample Size and Robust Estimation of Stochastic Frontier Production Models

Standardization and simplification of trade procedures constitute an urgent need to line up the document flow with the goods flow in international trade. Electronic data interchange (EDI) is a co‐operative inter‐organizational information system providing electronic exchange of messages agreed between trade partners. The paper outlines a possible implementation procedure for EDI. The first part consists of a six‐step feasibility study. If this study ends with a positive recommendation the implementation procedure can start. Discussions of this process aim to: completely understand about EDI, agree on standards with business partners, modify existing systems, translate data, prepare communications, and manage and audit the whole process. Concerning the cost‐benefit analysis for the EDI application—a step which is mandatory in the feasibility study—further details are given on different types of cost (recurring and non‐recurring, direct and indirect) and on both quantifiable and non‐quantifiable benefits.