Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WB73

4 - A Framework for Analyzing the U. S. Coin Supply Chain Yiwei Huang, Assistant Professor, Pennsylvania State University - Shenango, 173 Broadstone Drive, Mars, PA, 16046, United States, Subodha Kumar, Bala Shetty, Chelliah Sriskandarajah, Yunxia Zhu We analyze the supply side of the U.S. Coin Supply Chain with the objective of providing a near-optimal or an optimal operating policy that minimizes the total cost of producing, supplying, and managing coin inventory in the U.S. Coin Supply Chain. We develop efficient algorithms to solve various versions of the coin-management problem and perform an extensive analysis to answer several managerially relevant questions in the context of improving the efficiency of the U.S. Coin Supply Chain. n WB73 West Bldg 211B Practice- Optimization II Contributed Session Chair: Elham Taghizadeh, Wayne State University, Detroit, MI, 48035, United States 1 - Exact Penalization for Generalized Nash Equilibrium Problem Qin Ba, University of Southern California, Los Angeles, CA, United States, Jong-Shi Pang The Generalized Nash Equilibrium Problem (GNEP) extends the classical Nash Equilibrium Problem (NEP) by allowing individual players’ constraints, in addition to objectives, to depend on rivals’ decisions. It is an important model actively used in many different fields. However, solution algorithms are extremely scarce due to the presence of coupling constraints. This paper studies penalty methods which penalize violation of individual players’ coupling constraints and transform a GNEP into a single NEP. Several sufficient conditions are provided which guarantee exact penalization, i.e., the penalized NEP has identical solution sets as the original GNEP for finite penalty parameter. 2 - Underestimate Sequences via Quadratic Averaging Chenxin Ma, JD.com, Mountain View, CA, United States, Naga Venkata C. Gudapati, Majid Jahani, Rachael Tappenden, Martin Takác In this work we introduce the concept of an Underestimate Sequence. Our definition of a UES utilizes three sequences, one of which is a lower bound (or under-estimator) of the objective function. We propose several first order methods for minimizing strongly convex functions in both the smooth and composite cases. The algorithm have natural stopping conditions, which provides the user with a certificate of optimality. Convergence of all algorithms is guaranteed through the UES framework, and we show that all presented algorithms converge linearly, with the accelerated variants enjoying the optimal linear rate of convergence. 3 - Optimizing Schedule of Trains in Context of a Large Railway Network Srinivasa Prasanna, IIIT Bangalore, 26/C, Hosur Road, Electronics City, Opposite Infosys Technologies, Bangalore, 560100, India, Sanat Ramesh, Tarun Dutt, Anushka Chandrababu, Abhilasha Aswal We present two heuristics based on MILP formulations to optimize utilization for train timetabling problems for portions of one of the largest railway networks. These methods (extending state-of-art solvers) provide flexibility in scheduling additional trains while respecting a large number of constraints. We present methods to validate the deterministic schedule over global correlated variations in travel times without making any probabilistic assumptions. 4 - A Branch and Bound Algorithm for Two-machine No-wait Flow Shop Scheduling with Truncated Learning Function Vahid Azizi, Iowa State University, Ames, IA, United States, Guiping Hu There have been increasing interests in production scheduling considering learning effects. However, this problem has not been studied in a no-wait flow shop scheduling setting. This paper addresses a two-machine no-wait flow shop with the effect of the truncated learning function. With the truncated learning function, processing times of the jobs depend on their positions in the sequence and the learning parameter. A branch and bound algorithm is designed to minimize the makespan. A lower bound and two dominance properties are proposed which increase computational efficiency for the branch and bound algorithm.

5 - Global Non-probabilistic Validation of Schedules GN Srinivasa Prasanna, International Institute of Information Technology Bangalore (IIITB), Bengaluru, 560100, India, Anushka Chandrababu, Abhilasha Aswal, Sanat R, Tarun Dutt Deterministic optimization problems for train timetabling over even small portions of one of the world’s largest railway networks become intractable when uncertainty is introduced. We present methods based on linear programming to validate nominal schedules over global correlated variations in travel times (satisfying linear constraints) without making any probabilistic assumptions. 6 - A Hybrid Bat Algorithm for a Risk Averse Two Stage Stochastic Replenishment Problem with Transportation Costs Elham Taghizadeh, Wayne State University, Detriot, MI, 48202, United States, Saravanan Venkatachalam, Ratna Babu Chinnam Integrating inventory and transportation decisions can provide significant gains in the supply chain management. In this talk, we present a two-stage risk-averse stochastic programming model with Conditional-Value-at-Risk (CVaR) as risk- measure for a multi-item replenishment problem with transportation cost and demand uncertainty. To circumvent computational complexity, we develop a Hybrid Bat algorithm to solve the large-scale instances. Computational results based on sample average approximation approach will be presented. n WB75 West Bldg 212B Stakeholder Analysis Sponsored: Military and Security Sponsored Session Chair: Randy K. Buchanan, USACE - ERDC, Vicksburg, MS, 39180, United States Co-Chair: Simon Goerger 1 - Using Data Analytics on Stakeholder Requirements Data to Inform Research Innovation Christina Rinaudo, USACE Engineer Research and Development Center, 3909 Halls Ferry Road, Vicksburg, MS, 39180-6199, United States, Niki C. Goerger, Simon Goerger A recent stakeholder workshop to identify and discover challenges generated an initial data set of potential stakeholder requirements. Leveraging text and data analysis tools such as Tableau and R to analyze the stakeholder requirements data, areas for R&D efforts are illuminated to inform research innovation. 2 - Developing an Analytical Tool to Assess Stakeholder Engagements Janice P. Buchanan, USACE - ERDC, 3909 Halls Ferry Road, Vicksburg, MS, 39180, United States, Randy K. Buchanan, Simon Goerger, John R. Burt Identifying key stakeholders and tracking meeting results for a complex organization can be a daunting task. This presentation reviews the methodologies behind and development of StEAM - Stakeholder Engagement and Assessment Tool. StEAM is a prototype tool developed and used by the US Army Research and Development Center to help schedule stakeholder meetings, identify goals, and record accomplishments of senior leader engagements. 3 - Shared Vision Planning for Water Security – A Stakeholder Analysis Framework James Schreiner, U.S. Military Academy-West Point, Dept. of Systems Eng, Room 420, Mahan Hall, West Point, NY, 10996, United States Regional and Global water security represents an increasing risk and can be negatively or positively impacted by quality decisions of water management and infrastructure investment. Balancing watershed attributes of water storage, hydropower yield, risk mitigation, crop yield, and navigation among others becomes a difficult proposition when multi-national stakeholders are involved. This presentation will present a framework developed to use visual tools to capture stakeholder needs, wants, and desires while creating a shared situational awareness of the watershed system. The Tigris-Euphrates watershed provide the case study for this work.

468