INFORMS Nashville – 2016
428
2 - Incremental Methods For Additive Convex Cost Optimization
Mert Gurbuzbalaban, Postdoctoral Researcher, MIT,
176 Elm St #1, Building 32, Cambridge, MA, 02140, United States,
mert.gurbuzbalaban@gmail.com,Asuman Ozdaglar, Pablo Parrilo
Motivated by machine learning problems over large data sets and distributed
optimization over networks, we consider the problem of minimizing the sum of a
large number of convex functions. We develop and study incremental methods
for solving such problems, in particular for the random reshuffling method we
provide a sharp convergence result that answers an open question.
3 - Virtual Casted Supply Chain For Information Sharing
Mahamaya Mohanty, Research Scholar, IITDelhi,
Shaheed Jeet Singh Marg, New Delhi, 110016, India,
mahamayamohanty@gmail.com,Ravi Shankar
Cloud computing creates a multiplier effect in IT, supply in service chain
management with an intersection of virtualization and cloud where both fit
together. The cloud is a virtualization of resources that maintains and manages
itself as well as plays a vital role in making the overall system cost-effective,
enhanced flexibility, and sharing information with availability of resources thus
improving agility and enhancing flexibility. This article focuses on designing a
virtualized environment that is used to address a variety of business goals aimed
at improving efficiency and reducing costs of operation and maintenance of
physical servers.
WC04
101D-MCC
Power System Operation Models for Ramping
Scarcity Mitigation
Sponsored: Energy, Natural Res & the Environment,
Energy I Electricity
Sponsored Session
Chair: Masood Parvania, University of Utah, 50 S. Central Campus
Drive, Salt Lake City, UT, 84112, United States,
masood.parvania@utah.edu1 - Ramping Deliverability Enhancement Through
Flexible Transmission
Mostafa Ardakani-Sahraei, University of Utah, Salt Lake City, UT,
United States,
mostafa.ardakani@utah.eduProxy reserve constraints are employed within electricity market models in an
attempt to attain reliability in the face of uncertain load and element failures.
However, due to the complexities of the transmission network, the available
reserve, and specifically ramping capabilities, may not be deliverable to the
desired location. This talk discusses how flexibilities of the transmission network
can be exploited to enhance ramping deliverability. The results confirm the
effectiveness of the proposed method in improving reliability and reducing
reliability costs.
2 - Explicit And Implicit Mechanisms For Ensuring Reserve And
Ramping Capability
Erik Ela, EPRI,
eela@epri.comIn this presentation, we will present the central needs for operating reserve which
are used to meet variability and uncertainty. We compare the performance of
explicitly scheduled reserve, which is done through reserve inequality constraints
to meet the associated impact, with the implicitly scheduled operating reserve,
that in which the reserve is scheduled inherently by advanced scheduling
applications. We review case studies to compare performance in terms of
efficiently procuring sufficient capacity and ramp capability.
3 - Definition And Valuation Of Continuous-time Ramping Trajectory
In Power Systems Operation
Masood Parvania, University of Utah,
masood.parvania@utah.eduThe current discrete-time power system operation models imply that generating
units shall follow piecewise constant generation trajectories. This means that the
units’ ramping is modeled as the finite difference between the consecutive
generation samples. The discrete-time generation schedules and the resulting
rampings do not fully utilize the units’ flexibility to compensate the faster
variations of net-load, which may lead to ramping scarcity events. In this
presentation, we introduce a novel unit commitment model that schedules for
continuous-time generation and ramping trajectories, opening the door for
continuous-time valuation of ramping trajectories in power system operation.
4 - Wind Ramping Product For Power System Ramping Scarcity
Mitigation
Venkat Krishnan, NREL,
Venkat.Krishnan@nrel.govGolden, CO, 80401, United States,
Venkat.Krishnan@nrel.gov,Bri-Mathias Hodge, Anthony Florita
This talk will investigate the potential of wind power to provide ramping service
and the importance of ramp event forecasting. Forecasting mid-term (day-ahead
and intra-day) and short-term wind ramp events efficiently will be the basis of
managing and dispatching wind in a co-optimized energy and ramp service
markets. There are two aspects to this: 1) efficient wind forecasting platforms-
developed using big-data information processing technologies; and 2) wind ramp
event forecasting algorithm- based on optimized swinging door algorithm
(OpSDA) and dynamic programming.
WC05
101E-MCC
Spatial optimization
Sponsored: Energy, Natural Res & the Environment II Forestry
Sponsored Session
Chair: Nahid Jafari, University of Florida, Gainesville, FL, United States,
jnahid@hotmail.com1 - Fine-scale Spatial Targeting Of Surveillance To Minimize Costs Of
Invasive Species Introduction Across Large Landscapes
Rebecca Epanchin-Niell, Resources for the Future,
Epanchin-Niell@rff.orgEarly detection of new invasive species populations can reduce costs and damages
by allowing control when populations are smaller and less costly. This talk
presents a bioeconomic modeling approach to optimally target invasive species
surveillance at fine spatial scales (1 km2) to minimize total costs from invasions
and surveillance. The model accounts for spatial-temporal interdependencies
arising from invasion spread and surveillance costs. We apply the model to
surveillance for gypsy moth in the northwestern US.
2 - Analyzing Trade-offs Between Fire Prevention And Suppression In
The Republic Of Korea
Yohan Lee, Yeungnam University,
yohanlee76@gmail.comThis study explores the spatial tradeoff between the number of initial attack
firefighting resources and the level of fire prevention efforts mitigating the
probability of human-made fires in the Republic of Korea. To examine the spatial
trade-off, we utilize a hybrid system that combines a scenario-based, standard-
response optimization model with a stochastic simulation model. A mixed policy
that includes fire suppression and fire prevention efforts works better than a
single dominant policy such as a strong fire suppression policy, in particular, with
the consideration of spatial allocations because returns to effort in fire prevention
policy is dependent on the location of fires.
3 - Spatial Control Of The Argentine Black And White Tegu.
An Approach Using Linear Programming
Julien Martin, United States Geological Survey,
7920 NW 71 Street, Gainesville, FL, 32653, United States,
julienmartin@usgs.gov,Mathieu Bonneau, Fred A. Johnson,
Brian Smith, Christina M. Romagosa
We propose to frame spatial control of tegu in south Florida as a linear
programming optimization problem. This formulation with a discrete reaction-
diffusion model permits calculation of an optimal control strategy that minimizes
the remaining number of tegus for a fixed cost or that minimizes the control cost
to achieve containment. We compute the optimal strategy for a range of possible
model parameters and discuss the best strategy to use in practice as a function of
the risk attitude of the decision maker.
4 - Robust Spatial Optimization For The Invasive
Species Management
Nahid Jafari, University of Florida,
nahid.jafari@ufl.eduThe problem of invasive species management concerns modeling the pattern of
spread of the invasive, estimation of control costs, spatial design of the control
effort, and accounting for uncertainties in model parameters. Robust optimization
constructs a solution that is feasible for any realization of the uncertainty in a
given uncertainty set (achieves the best worst-case objective function value).
Given the computational efficiency of robust optimization, we are developing a
spatial-optimization model to select sites for efficiently controlling invasive species
to minimize their ecological damage, as well as to minimize the costs given
limited financial resources.
WC04