Issue |
RAIRO-Oper. Res.
Volume 48, Number 3, July-September 2014
|
|
---|---|---|
Page(s) | 303 - 324 | |
DOI | https://doi.org/10.1051/ro/2014009 | |
Published online | 21 March 2014 |
Process parameter prediction via markov models of sub-activities
1 Department of Industrial Engineering, POLI, Federal
University of Rio de Janeiro, CP 64548, 21941-972 Rio de Janeiro, Brazil.
lgmarujo@ufrj.br
2 Department of Ocean Engineering,
COPPE, Federal University of Rio de Janeiro, CP 68508, 21941-972
Rio de Janeiro,
Brazil
Received:
21
August
2013
Accepted:
6
January
2014
This work aims to fill a lacunae in the project-oriented production systems literature providing a formal analytic description of the rework effects formulae and the determination of the extended design time due to a certain degree of overlapping in a pair of activities. It is made through the utilization of concepts of workflow construction with hidden (semi) Markov models theory and establishing a way to disaggregate activities into sub-activities, in order to determine the activity parameters used by the project scheduling techniques. With the aim to make a correlation between the entropy of the state transitions and the probability of changes, the information theory is also used, and the concept of impact caused by the probability of changes is provided. Numerical examples are shown for the purpose to demonstrate the applicability of the concepts developed, and one example of overlapping of two activities is shown. The original contributions of this work are shown on the last section.
Mathematics Subject Classification: 90B15 / 90B30 / 68M20
Key words: Activity parameters / sub-activities Markov model / entropy / project scheduling parameters / rework estimation
© EDP Sciences, ROADEF, SMAI, 2014
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.