26.11.2012 Views

Technical Sessions – Monday July 11

Technical Sessions – Monday July 11

Technical Sessions – Monday July 11

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

MB-08 IFORS 20<strong>11</strong> - Melbourne<br />

The placement of a (possibly nonconvex) shape into a container, via translation,<br />

is a fundamental operation useful in containment, packing and layout problems.<br />

Using the Minkowski sum here requires a complement operator with<br />

unbounded sets that can be handled by various limited methods. We discuss<br />

an alternative formulation that eliminates unbounded sets. This is motivated by<br />

work on packing 3D suitcases into a car trunk. We prove that this yields the<br />

same result as the standard translational single-item containment algorithm,<br />

with some restrictions, and that it works in arbitrary dimensions.<br />

3 - A Priority-considering Approach for the Threedimensional<br />

Bin Packing Problem<br />

Jidong Ren, Graduate School of Engineering, Kyoto University,<br />

Yoshida Honmachi, Sakyo-ku, 6068501, Kyoto, Japan,<br />

renjd<strong>11</strong>1@hotmail.com, Yajie Tian, Tetsuo Sawaragi<br />

A heuristic approach is proposed to solve the three-dimensional bin packing<br />

problem in which a given set of three-dimensional rectangular items should<br />

be packed into the minimum number of identical finite containers. The core<br />

idea of this approach is that some items with large volume are preferentially<br />

assigned into the containers. Within the approach a single container algorithm<br />

is used which is generalized from the algorithm proposed in Ren et al. (20<strong>11</strong>).<br />

The proposed approach achieves excellent results for the test cases suggested<br />

by Ivancic et al. (1989) with reasonable computing time.<br />

4 - Multi-sphere Scheme with a General Nonlinear Programming<br />

Solver<br />

Takashi Imamichi, IBM Research - Tokyo, 1623-14,<br />

Shimotsuruma, 242-8502, Yamato, Kanagawa, Japan,<br />

imamichi@jp.ibm.com, Ernesto G. Birgin<br />

The multi-sphere scheme is a general framework for finding a layout of given<br />

shapes into a container. It first approximates each shape by a set of circles or<br />

spheres and then find a layout of the sets of circles or spheres. The author applied<br />

the quasi-Newton method, which is for unconstrained nonlinear programming<br />

problems, to finding a layout in the previous work. In this talk, we apply<br />

a general nonlinear programming solver ‘ALGENCAN’ instead, construct an<br />

algorithm for the two-dimensional strip packing problem with free-rotations,<br />

and report the results of experiments.<br />

� MB-08<br />

<strong>Monday</strong>, 14:00-15:30<br />

Meeting Room 107<br />

Algorithms for Nucleic Acid Analysis<br />

Stream: Bioinformatics<br />

Invited session<br />

Chair: Marta Szachniuk, Institute of Bioorganic Chemistry, PAS,<br />

Noskowskiego 12/14, 61-704, Poznan, Poland,<br />

Marta.Szachniuk@cs.put.poznan.pl<br />

1 - RNA Structure Prediction from OR Perspective<br />

Marta Szachniuk, Institute of Bioorganic Chemistry, PAS,<br />

Noskowskiego 12/14, 61-704, Poznan, Poland,<br />

Marta.Szachniuk@cs.put.poznan.pl<br />

RNA molecules serve diverse structural, catalytic and regulatory function in<br />

living cells. The knowledge of their 3D folds is essential to understand the increasing<br />

number of their biological functions. The talk will present a problem<br />

of RNA structure modeling from the OR perspective. A new approach to the<br />

RNA comparative modeling based on fragment matching and assembly will<br />

be highlighted. We will focus on solutions applied on the different levels of<br />

structure composition process, like graph theory based approach to secondary<br />

structure fragmentation, module alignment and merging.<br />

2 - Match-searching Algorithms for Gene Silencing Assessment<br />

Mark Horn, Mathematics, Informatics & Statistics, CSIRO,<br />

Locked Bag 17, 1670, North Ryde, NSW, Australia,<br />

Mark.Horn@csiro.au<br />

Gene silencing applications involve the use of synthetic RNA silencer<br />

molecules with sequences designed to suppress specific genes. Reliable assessment<br />

of silencer impacts requires the rapid identification of matching sites<br />

in messenger RNA. The paper presents algorithms and data structures for this<br />

task, based on a compact numerical encoding of sequence data. The main algorithm<br />

involves a search for a Boolean dominance relation, which is related<br />

to a database task called skyline search. Computational tests show that the<br />

algorithm is significantly faster than conventional alternatives.<br />

14<br />

3 - Inversion Distributions in RNA and Their Roles in Secondary<br />

Structure Prediction<br />

Ming-Ying Leung, The University of Texas at El Paso, TX<br />

79968-0514, El Paso, United States, mleung@utep.edu, Sameera<br />

Viswakula, Michela Taufer, Kyle Johnson<br />

RNA secondary structures (SS) provide information for understanding replication<br />

mechanisms of viruses with critical influence on agricultural resource<br />

sustainability. As sequence lengths of viral genomic RNA often exceed limits<br />

allowed by SS prediction software, we develop strategies to cut long RNA sequences<br />

into shorter chunks to be predicted separately. Individual predictions<br />

are then assembled to form an overall SS. Statistical distribution properties of<br />

inversions, which are essential for all SS, can help establish criteria in the segmentation<br />

process to maximize prediction accuracy.<br />

� MB-09<br />

<strong>Monday</strong>, 14:00-15:30<br />

Meeting Room 108<br />

Rich Vehicle Routing<br />

Stream: Vehicle Routing<br />

Invited session<br />

Chair: Marc Sevaux, Lab-STICC - CNRS UMR 3192, Université de<br />

Bretagne Sud - UEB, Centre de Recherche - BP 92<strong>11</strong>6, 2 rue de Saint<br />

Maude, 56321, Lorient, France, marc.sevaux@univ-ubs.fr<br />

Chair: Kenneth Sörensen, Faculteit Toegepaste Economische<br />

Wetenschappen, Universiteit Antwerpen, Prinsstraat 13, 2000,<br />

Antwerpen, Belgium, kenneth.sorensen@ua.ac.be<br />

Chair: Richard Hartl, Dept of Management, University of Vienna,<br />

Bruenner Str. 72, A-1210, Vienna, Austria, richard.hartl@univie.ac.at<br />

1 - Fuel Efficient Haul Truck Routing on Mines<br />

Kevin Duffy, Institute of Systems Science, Durban University of<br />

Technology, Berea Rd, 4000, Durban, Kzn, South Africa,<br />

kevind@dut.ac.za<br />

This research investigates computational methods for estimating fuel efficient<br />

path calculations for a mine road network based on fuel consumed by large<br />

haul trucks. Algorithms using both shortest path and fuel efficient path computations<br />

are compared. Depending on road condition the fuel efficient path<br />

can save more fuel. The results are discussed with reference to extreme and<br />

general situations. By scaling the results up to realistic mine proportions it is<br />

shown how significant saving in production costs should be possible where a<br />

fuel efficient haul road utilization system is implemented.<br />

2 - A Tabu Search Method for the Open Vehicle Routing<br />

Problem with Time Windows<br />

Leon Li, Department of Logistics & Maritime Studies, The Hong<br />

Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong,<br />

lgtleon@polyu.edu.hk, Zhuo Fu, Richard Eglese<br />

In this paper, a tabu search heuristic with an intensive search mechanism for<br />

the open vehicle routing problem with soft time windows (OVRPSTW) is proposed,<br />

by which the problem with hard time windows (OVRPHTW) can also be<br />

solved by simply setting appropriate parameters in the penalty function. Computational<br />

results on the benchmark problems are provided and compared with<br />

results in the literature, which show that it can produce good solutions. Comparisons<br />

of the features of two different soft time window constraint types are<br />

also carried out.<br />

3 - A New Solution Approach for the Rollon-rolloff Vehicle<br />

Routing Problem<br />

Juyoung Wy, Industrial and Management Engineering, Pohang<br />

University of Science & Technology, 790-784, Pohang, Korea,<br />

Republic Of, hwiriric@postech.ac.kr, Byung-In Kim<br />

The rollon-rolloff vehicle routing problem, in which large garbage collection<br />

containers are at customer locations such as construction sites and shopping<br />

centers, is handled. Tractors transport a container at a time between locations.<br />

We propose an integrated heuristic algorithm which can construct initial solutions<br />

and improve the solutions iteratively. The construction algorithm is based<br />

on the Sweep Nearest Algorithm and the improvement algorithm includes perturbation,<br />

inter- and intra-route algorithms. New best solutions for the benchmark<br />

problems were found using the proposed approach.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!