2016 INFORMS Annual Meeting Program

397 / 561

INFORMS Nashville – 2016
397
4 - Implement Real Time Pricing With Regret Based Learning
Andrew Lu Liu, Purdue University,
The situation where price-responsive consumers determine what to do in the
near future (such as when to charge their PEVs) forms a dynamic and
incomplete-information game, in which the consumers’ collective actions will
impact electricity prices, which in turn affect their payoffs. We propose a regret-
matching-based algorithm for each consumer to learn their strategies, as opposed
to naively responding to day-ahead prices. We will show convergence to a
correlated equilibrium of the regret-matching approach, and study the price of
anarchy of the regret-based learning approach.
WB05
101E-MCC
Forest management: Transportation and/or
Collaborative Logistics
Sponsored: Energy, Natural Res & the Environment II Forestry
Sponsored Session
Chair: Mikael Ronnqvist, Universite Laval, Quebec, QC, Canada,
1 - Overlapping Coalitions In Collaborative Transportation Of
Forest Biomass
Mario Guajardo, Norwegian School of Economics,
, Patrik Flisberg, Mikael Frisk,
Mikael Ronnqvist
Given a set of companies, we address a collaborative transportation problem in
which a same company can collaborate in more than one coalition. This problem
corresponds to what in cooperative game theory is known as coalition
configuration. We develop area-driven approaches and another approach that
embeds the coalition configuration in the transportation problem. We report
numerical results for a case in Sweden involving about 6 million tons of forest
biomass. Collaboration in this case has not only the potential of cost savings, but
also of increasing the use of bioenergy. The results improve substantially
compared to a coalition structure and are competitive with the grand coalition.
2 - Bundle-based Auction For Timber Allocations
Sophie D Amours, Université Laval,
Marc-André Carle,
Mikael Ronnqvist
Tenure of the public forest in Québec, Canada, includes the usage of an
auctioning system to allocate the 25% of the commercial forest to industry. We
have studied the system and explored new strategies to increase its performance.
This includes using combinatorial auctions. The presentation will review the
context, the literature and illustrate using a case study the potential and
challenges of implementing a combinatorial auctioning system. These include
higher efficiency in coordinating needs with supplies as well as the potential
challenge of finding multiple optimal solutions and the need for discriminating
criteria.
3 - Supply Network Planning Of Forest Fuel Using
Robust Optimization
Jens Bengtsson, Norwegian University of Life Sciences,
, Patrik Flisberg, Mikael Ronnqvist
Fluctuating and uncertain temperature during the year will cause an uncertain
demand at heating plants. These plants use district heating with an energy supply
of forest fuel. Changed supply contracts with customers (heating plants) on a
yearly frequency may result in very different optimal supply network. We use
robust optimization to analyze where forests fuel storage points should be located
and their capacities in order to take transportation cost and lead times into
account. Critical is to ensure that there is enough supply and transport capacity
available throughout the year.
4 - Collaborative Logistics
Mikael Ronnqvist, Universite Laval,
Collaborative logistics is an efficient approach to improve the logistics operations.
In our presentation, we consider horizontal collaboration between several
companies and larges savings have been reported in the literature. However, there
is also a number of practical issues to deal with including sharing principles and
how the joint planning is done without revealing sensitive information. We
describe a number of practical applications and discuss their potentials and
pitfalls.
WB07
102B-MCC
Simulation and Data Mining
Sponsored: Data Mining
Sponsored Session
Chair: Mohammad Ali Asudegi, UTK, 301 Woodlawn Pike, Apt G12,
Knoxville, TN, 37920, United States,
1 - Discrete Event Simulations Of Urgent Care Resource Scheduling
Leyla Zhuhadar, Assistant Professor, Western Kentucky University,
1906 College Heights Boulevard, 226 Grise Hall, Bowling Green,
KY, 42101, United States, ,
Evelyn Thrasher, Kirk Atkinson
Discrete-event simulation is used to estimate the average amount of time that a
patient spends waiting to see a nurse and/or waiting to see a doctor. In addition,
the model has been extended to estimate the effect on the waiting times if
additional nurses or doctors are hired. In this presentation, a study of the value of
multiple replications for discrete-event simulation models is discussed and factors
to enable greater control on multiple design points with the experiment are
defined. Finally, to control the simulation runs for easier model verification,
random seeds are used.
2 - A Data Mining Approach For Evaluating Simulation Output
Thomas Robbins, East Carolina University, 3212 Bate Building,
Greenville, NC, 27858, United States,
We conduct a detailed simulation experiment to compare two different queuing
models over a wide range of inputs with the goal of determining the conditions
under which each model works best. We implement a data mining approach to
characterize the outputs and evaluate the input conditions that best predict model
superiority. Since performance can be evaluated across multiple measures we use
a clustering approach to characterize output performance. We then investigate
multiple classification techniques to isolate inputs that best predict model
superiority.
3 - Performance Indicators And Operating Factors In A Complex
Manufacturing System
Mohammad Ali Asudegi, UTK, 525 John D. Tickle Engineering
Building, 851 Neyland Dr, Knoxville, TN, 37996, United States,
Rupy Sawhney
Reliability, Lead-Time and Cycle-Time are the first and most important key
performance indicators, KPIs, of each manufacturing system. Understanding the
relation between operating factors, system design factors and key performance
indicators would help to be able to make better scheduling and run the system
within its optimal specification.A simulation has been conducted to study the
relation between the operating speed of each machine in a complex system with
Lead-Time, Cycle-Time and the reliability of the whole system.
WB08
103A-MCC
Supply Chain Mgt, Green
Contributed Session
Chair: Jianan Sun, Doctoral Student, Xi’an Jiao Tong University,
28 Xianning Road, Shanxi Province, Xi an, 710049, China,
1 - Price Competition Induced By A Partially Green-sensitive
Consumer Market In A Dual Manufacturer-single Retailer Channel
Arka Mukherjee, PhD Scholar, Concordia University,
1455 De Maisonneuve Blvd. W., Montreal, QC, H2G 1M8, Canada,
In a marketing channel with two manufacturers and one retailer, price
competition between green and non-green products is studied. Considering
different market power structures Game theoretic models are developed.A simple
Nash game, a cost-sharing contract and a scenario where manufacturers are the
market leaders are considered. The results from the different power structures are
compared analytically and numerically and interesting conclusions are drawn.
2 - Incentives For Fair Labor Practices In a Supply Chain
Susan A Slotnick, Professor, Cleveland State University, 1860 E.
18th Street, BU 542, Cleveland, OH, 44115, United States,
Matthew J Sobel
This paper investigates the tradeoffs involved in a firm’s decision to implement
policies and procedures to ensure that its products are not manufactured using
unfair labor practices. We consider a firm that repeatedly sources a product from a
single supplier, maximizes its own profit, and decides whether to cut its profit
margins and/or expend resources to inspect the supplier, as well as to improve
and/or maintain fair labor practices at that supplier. Retail demand for the
product is influenced by reputation.
WB08