nurse_rostering
所属分类:医药行业
开发工具:Visual C++
文件大小:1010KB
下载次数:22
上传日期:2012-10-01 16:46:16
上 传 者:
suntoward
说明: 护士排班问题是个NP-hard问题,用LocalSover算子可以很好解决,该算子需要另外安装,以用以学习目的。
(nurse rostering as a NP-hard problem using with LocalSolver.com s LocalSolver slover which can be used freely on condition for learning)
文件列表:
nurse_rostering (0, 2012-09-26)
nurse_rostering\evaluator (0, 2012-09-26)
nurse_rostering\evaluator\evaluator.jar (998840, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden01.xml (28747, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden02.xml (28747, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden03.xml (39589, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden04.xml (31439, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden05.xml (31398, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden06.xml (46510, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden07.xml (46509, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden08.xml (57351, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden09.xml (48928, 2012-07-31)
nurse_rostering\evaluator\sprint_hidden10.xml (49203, 2012-07-31)
nurse_rostering\instances (0, 2012-09-26)
nurse_rostering\instances\sprint_hidden01.txt (6313, 2012-07-31)
nurse_rostering\instances\sprint_hidden02.txt (6313, 2012-07-31)
nurse_rostering\instances\sprint_hidden03.txt (7996, 2012-07-31)
nurse_rostering\instances\sprint_hidden04.txt (6556, 2012-07-31)
nurse_rostering\instances\sprint_hidden05.txt (6607, 2012-07-31)
nurse_rostering\instances\sprint_hidden06.txt (9324, 2012-07-31)
nurse_rostering\instances\sprint_hidden07.txt (9323, 2012-07-31)
nurse_rostering\instances\sprint_hidden08.txt (11006, 2012-07-31)
nurse_rostering\instances\sprint_hidden09.txt (9566, 2012-07-31)
nurse_rostering\instances\sprint_hidden10.txt (9618, 2012-07-31)
nurse_rostering\nurse_rostering.lsp (17960, 2012-08-01)
NURSE ROSTERING (PATAT'10 competition)
''A contribution of B. Dessort, C. Groix, C. Jaulin , Ecole des Mines de Nantes (France)''
The nurse rostering problem involves assigning shifts to nurses taking several
constraints into account. Here the only hard constraints are that each nurse
must be assigned exactly one shift per day and that daily coverage constraints
must be satisfied every day. The objective function is a linear combination of
penalties associated to the satisfaction of a certain number of scheduling rules:
min/max number of working days, min/ max number of consecutive working days,
identical shifts during week-ends, unwanted shift patterns...
This problem was the subject of the PATAT 2010 Nurse Rostering Competition. More
information can be obtained on [[http://www.kuleuven-kulak.be/nrpcompetition]].
Text data files from the final stage of this competition are available in the
instances/ folder. More can be downloaded from the competition website. The
program can be launched with the following command:
localsolver nurse_rostering.lsp inFileName=instances/sprint_hidden01.txt solFileName=solution01.xml lsTimeLimit=10
It launches localsolver for 10 seconds on this model. The output file in written
in xml format and can be checked with the official evaluator of the competition.
For some reason, this evaluator requires the XML input file, hence these XML inputs
are provided in the evaluator folder. The following command checks the validity
of the solution computed above and compute its cost.
java -jar evaluator/evaluator.jar evaluator/sprint_hidden01.xml solution01.xml
From a modeling perspective, the model is classically based on X[e,d,s] decision
variables, equal to 1 when employee e is assigned to shift s on day d. Remember
that LocalSolver offers many non linear operators including conditional operators.
For instance the penalties on consecutive working days are modeled with intermediate
variables nbSlidingConsecutive[e][d] equal to the number of identical days
(worked or off) preceeding day d in the schedule of employee e. Each such
nbSlidingConsecutive[e][d] variable is defined recursively from variable
nbSlidingConsecutive[e][d-1] with a conditional operator. Finally, detecting too
long or too short sequences is straightforward.
For more details, see [[http://www.kuleuven-kulak.be/nrpcompetition]].
近期下载者:
相关文件:
收藏者: