Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

WA34

2 - Optimal Cooperative Stocking Policy under Demand and Lead-time Uncertainties with Space Constraint and Penalty for Late Delivery Md Shahriar Jahan Hossain, Louisiana State University, 4141 Burbank Dr. Apt. 3, Baton Rouge, LA, 70808, United States, Bhaba R. Sarker This research presents a single-vendor single-buyer cooperative inventory model for a nonperishable item. The model deals with both demand and lead-time uncertainties, space constraint and a penalty cost for late delivery. A joint total cost function is formulated as a constrained non-linear programming problem. The joint total cost is minimized in order to determine the optimum order quantity, reorder point and number of shipments. Solution procedures are demonstrated through complete numerical examples with different probability distributions of demand and lead time. A set of efficient and good solutions obtained are predictive of an economic operating policy for the joint contract. 3 - Strategic Safety Placement under Stochasticoptimal Control Policy Strategic safety stock placement problem is a tactical model to determine optimal places and levels of safety stocks in multi-echelon system. Existing models, such as guaranteed-service model or stochastic-service model, use the base-stock policy as a building block, under which each stage observes demand and places an order on its suppliers equal to the observed demand. In this study, we propose a safety stock placement model under stochastic optimal control policy, where order- quantity is calculated so as to minimize weighted sum of variation of inventory and order-quantity over infinite time horizon. By doing so, variation of order- quantity can be reduced as well as inventory level. n WA34 North Bldg 223 Practice- Disaster and Disruption Management II Contributed Session Chair: John Doerpinghaus, University of Arkansas, Fayetteville, AR, 72701, United States 1 - An Optimization Model for Wildfire Suppression Process and Resident Evacuation Siqiong Zhou, San Jose State University, San Jose, CA, 95123, United States, Ayca Erdogan Wildland-urban interface (WUI) wildfires have been big threats in many countries. We present a two-stage stochastic integer programming model for efficient wildfire comprehensive response on firefighting resource allocation and resident evacuation. This model minimizes the number of residents at risk and considers the total cost of resource operation and property loss. Resource preparation and allocation decisions are made based on both fire spread process and population density. 2 - Smart City Emergency Response Simulation Framework Charles Njelita, Data Scientist, TATA America International Inc, 379 Thornall Street, Edison, NJ, 08837, United States Cities around the world are becoming smarter and one of the criteria of a smart city is its ability to evacuate its citizens and visitors in the case of natural disaster, bomb blast, highway accident etc. How fast the first responder to the scene of the incidence determines the number of lives saved and injury prevented. The purpose of the paper is to develop a simulation framework which can be used by decision makers in response to an emergency that may have impacts on multiple aspects of the urban environments and socio-economical operations. These researchers will develop a multi-level simulation framework used by decision makers in city to make timely and effective responses to an emergency. 3 - Simulating and Analyzing Cascading Failures in Power Networks in Disaster Aftermath Alireza Inanlouganji, Arizona State University, 699 S. Mill Ave, Tempe, AZ, 85281, United States, Giulia Pedrielli Proactive response to disasters is an increasingly interesting area of research with high potential of cost reduction and safety improvement. We propose a novel framework to efficiently simulate and quickly analyze power networks and their interactions with other critical interdependent infrastructures. Both initial, exogenous, failures and failures due to cascading effects are considered and we seek a set of interesting scenarios that lead to worst adverse effects on the power network using an efficient simulation-optimization algorithm. Shunichi Ohmori, Assitant Professor, Waseda University, Room 0903A, Okubo 3-4-1, Shinjuku, Tokyo, Japan

4 - Management of Interdependent Infrastructure Networks under Disaster-related Uncertainties Tugce Canbilen, Research Assistant, Middle East Technical University, Universiteler Mahallesi, Dumlupinar Bulvari No: 1, Endustri Myhendisligi Balumu, Ofis: 326, Ankara, 06800, Turkey, Sakine Batun, Melih Celik After a disaster, multiple infrastructures have disruptions in their services. During the restoration of these services, we need to consider the operational and restoration interdependencies among these infrastructures. In this study, the problem under consideration is Interdependent Infrastructure Restoration with disaster related uncertainties. We propose a two-stage program for this problem. First stage makes reinforcement decisions on the arcs of the network while second stage schedules the restoration activities considering the available resources and inherent interdependencies in the network. We will present our computational tests and preliminary results in the presentation. 5 - Classification of the Interdependencies in the Food and Agriculture Critical Infrastructure in Arkansas John Doerpinghaus, University of Arkansas, 4207 Bell Engineering Center, Fayetteville, AR, 72701, United States, Sarah G. Nurre, Kelly Sullivan, Benjamin R. Runkle Classifying infrastructure interdependencies is vital for the creation of realistic disaster preparedness and response models. The existing interdependency classifications do not capture many intricacies of the critical food and agriculture infrastructure. We propose new infrastructure interdependency classifications relevant to food and agriculture which can be generalized to many critical infrastructures. With each new classification, we provide real accounts and examples based on over 30 interviews with stakeholders in the food and agriculture critical infrastructure sector. n WA35 North Bldg 224A Passenger Demand Forecasting and Airline Capacity Management Sponsored: Aviation Applications Sponsored Session Chair: Farbod Farhadi, Roger Williams University, 1 Old Ferry Road, Bristol, RI, 02809, United States 1 - A Modeling Framework for the Airline Near-term Capacity Adjustment Problem Ahmed Abdelghany, Professor, Embry-Riddle Aeronautical University, College of Business, 600 S. Clyde Morris Boulevard, Daytona Beach, FL, 32114, United States, Khaled Abdelghany, Farshid Azadian A modeling framework for the airline’s near-term capacity adjustment problem is presented. The input to the model is an initial schedule for the host airline and any expected near-term changes in origin-destination (OD) demand, competition, and resources (aircraft, gates, etc.). The model searches for the best possible adjustments for flight frequency for each fleet type in each city-pair to maximize profitability. The performance of the framework is evaluated through several experiments in a hypothetical setting that mimics the U.S. domestic market, where the model is used to study the capacity adjustment for two major airlines. 2 - Flight Outsourcing in the U.S. Domestic Market David F. Pyke, University of San Diego, Coronado 108, 5998 Alcala Park, San Diego, CA, 92110-2492, United States, Farbod Farhadi, Soheil Sibdari Major airlines, while maintaining route and airport dominance, outsource a portion of their services to smaller operators for a fee. Smaller operators offer their services either exclusively to a single major airline, or partner with multiple airlines within the same markets. In this research, we investigate the impact of such phenomena on total capacity and average prices. 3 - Passengers Demand Forecasting of the U.S. Domestic Airline Market Nahid Jafari, Farmingdale State College, 2350 Broadhollow Road, Farmingdale, NY, 11735, United States This research examines the strategies followed by the U.S. domestic airlines in 21st century to achieve their highest revenue in regard with their capacity choices. During this period, although the airlines experienced higher load factors and inflated airfares, in addition to the fluctuation of fuel expenses, they were having profit losses. The analysis is conducted on 16 major airlines on primary factors such as the load factor, market supply (seats) and demand (passengers), unemployment rate, fuel expenses, flight frequency, and aircraft size. We deploy the Long Short-Term Memory network to forecast the U.S. domestic airline passengers.

422