2015 Informs Annual Meeting

WE77

INFORMS Philadelphia – 2015

3 - Optimal Maintenance Policy with Repair Time under Warranty Minjae Park, Hongik University, 72-1 Sangsu-Dong, Mapo-Gu, Business School, Seoul, 121-791, Korea, Republic of, mjpark@hongik.ac.kr This study formulates a warranty cost model for the repairable products when an age replacement policy is adopted in cooperation with the renewing minimal repair-replacement warranty and studies the optimal choice of the preventive replacement age. Under the renewing minimal repair-replacement warranty, either minimal repair or replacement is performed depending on the length of repair time when the product failures occur during the warranty period. WE76 76-Room 204C, CC Simulation and Optimization Applications Contributed Session Chair: Hossein Hashemi, Graduate Research Assistant, Southern Methodist University, 3101 Dyer Street, Dallas, TX, 75205, United States of America, shashemi@smu.edu 1 - Mathematical Analysis for Tomato Spotted Wilt Virus Transmission Yan Kuang, Kansas State University, Manhattan, KS, 66502, United States of America, ykuang@ksu.edu, Songinan Zhao, David Ben-arieh, Chih-hang Wu Tomato spotted wilt virus (TSWV) is a plant infecting virus and impacts many food and ornamental crops in both quantity and quality aspects, resulting in crop disease epidemics of worldwide economic significance. TSWV is transmitted particularly by thrips, the most efficient of which is West Flower Thrip (WFT). This research is aimed to use mathematical models to study the disease transmission effects. Through the preliminary results, we discuss the impacts of potential mitigation strategies. 2 - A Farmers Markets Location Allocation Framework for Enhanced Public Health Hoyoung Na, University of Arizona, 8131 N Midnight Way, Tucson, AZ, 85741, United States of America, nhy4201@email.arizona.edu, Sojung Kim, Young-jun Son, Langellier Brent The goal of this study is to propose a new location allocation framework of farmers markets (FMs) to enhance public health, by introducing FMs into food deserts which have a limited access to healthy food resources. The proposed framework adopts the theory of planned behavior (TPB) under an agent-based simulation (ABS) platform for food choice behavior modeling of individuals. The framework is demonstrated via AnyLogicÆ ABS software with the 2007 Food Attitudes and Behaviors Survey dataset. 3 - A Network Size-Reduction Methodology for Stochastic Prediction of Wildfire Propagation Mohammad Hajian, PhD, Northeastern University, Dept. of Mechanical and Industrial Eng., 360 Huntington Avenue, Boston, MA, 02115, United States of America, mhajian@coe.neu.edu, Peter Kubat, Emanuel Melachrinoudis Each year, thousands of wildfire events throughout the US threaten life and property. In a fire event, it is important to make the most effective decision under time pressure. This requires a fast and accurate prediction of the fire propagation, taking into account uncertain factors such as wind. In this research, we model the wildfire propagation as stochastic shortest path and present a network reduction methodology to effectively predict the wildfire traversal time. 4 - An Enhanced Global Optimization Method for a PV-Diesel-Battery Hybrid Power System Siew Fang Woon, Universiti Utara Malaysia, 06010 Sintok, Kedah, Malaysia, woonsiewfang@yahoo.com, Sie Long Kek, Muhammad Nazri Abu Bakar Non-linear constrained discrete-valued optimal control problems are known to have multiple local optima that require global optimization methods to find the best solution. We developed a global optimization method by embedding an improved discrete filled function technique into a computational optimal control algorithm. This proposed method improves the computational efficiency in determining a near global optimal solution for the operating cost of a PV-diesel- battery hybrid power system.

5 - Multi-agent Learning Approach for Online Calibration of Real-time Traffic Network Management System Hossein Hashemi, Graduate Research Assistant, Southern Methodist University, 3101 Dyer Street, Dallas, TX, 75205, United States of America, shashemi@smu.edu, Khaled Abdelghany The paper introduces a novel formulation and a solution methodology for the problem of online calibration of simulation-based Dynamic Traffic Assignment (DTA) models. The methodology calibrates the model considering multiple inconsistency sources including the time-dependent demand pattern and the traffic flow propagation models. It adopts a reinforcement learning approach to determine an efficient activation schedule for the different model adjustment modules. WE77 77-Room 300, CC Decision Analysis for Supply Chains Contributed Session Chair: Robert Inman, General Motors, MC 480-106-RE1, 3 0500 Mound Road, Warren, MI, 48090, United States of America, robert.inman@gm.com 1 - The Nexus Between GDP, Population Growth/fuel Prices in the United States Fesseha Gebremikael, PhD Student, UGPTI, P.Box 6050, Fargo, ND, 58102, United States of America, fesseha.gebremikael@ndsu.edu, Joseph Zmerekovsky, Karen Froelich, En Sue Lee The linkage between GDP and VMT, population growth, and fuel prices, has been strong for quite some time, irrespective of the ups and downs of the national economics of the developed societies, particularly the United States. Linear regression model will be used to test empirically the nexus between VMT and the related variables. Data will be collected from the World Bank, BEA, EIA and other relevant sources /compiled to gauge the change over time. 2 - An Integrated Model of Location and Safety Stock Placement Liwen Cui, PhD Candidate, Tsinghua University, Room 615, Shunde Building, Beijing, 100084, China, cuiliwen0512@126.com, Zuo-jun Max Shen We study the integrated location and safety stock placement problem, and design an efficient dynamic programming algorithm to solve the problem. From analytical analysis and numerical studies on real supply chain data, we obtain some interesting managerial insights. 3 - Performance Outcomes and Success Factors of Healthcare Industrial Vending Solutions Healthcare firms face considerable pressure to efficiently and effectively manage their inventory. Industrial vending machine solutions, a specific form of VMI, are one solution. IVM includes automated drug distribution systems and medical supply dispensing systems (e.g., Pyxis). Two-hundred nine healthcare supply chain managers were used to test the theoretical model. Results show complex interactions of the enablers with one another and in their relationship with perceived success. 4 - Ordering Problems with Demand Forecast Updating and Supply Constraints John Kros, Professor Of Marketing And Supply Chain Management, East Carolina University, 3205 Harold Bate Building, Greenville, NC, 28590, United States of America, krosj@ecu.edu This paper investigates an extension of the newsvendor model with demand forecast updating under supply constraints. The retailer prefers to postpone the order for improved demand accuracy. However, the postponement is associated with a price to the supplier and thus is limited.In this situation, we investigate the optimal ordering decisions based on demand forecast updating, increased purchasing cost, and restrictions on the ordering time and quantity. 5 - Product Complexity and Supply Chain Design Robert Inman, General Motors, MC 480-106-RE1, 3 0500 Mound Road, Warren, MI, 48090, United States of America, robert.inman@gm.com, Dennis Blumenfeld Assembling a product requires each and every part. Compared to assemblers of simple products, manufacturers of complex products are much more sensitive to supply chain delays. This heightened vulnerability to supply chain disruptions should lead complex product assemblers to design less risky supply chains. This paper models how product complexity drives the likelihood of production disruption and provides insights for supply chain design. Meimei Zheng, Nanyang Technological University, 50 Nanyang Avenue, Singapore, 639798, Singapore, meimeizheng2009@gmail.com, Kan Wu

503