Informs Annual Meeting 2017

TD56

INFORMS Houston – 2017

3 - Limited Foresight Strategy in Multi-period Mean-variance Framework

4 - Heuristic Algorithm for Megacity Vehicle Routing Problem Jihyun Jo, Pennsylvania State University, 234 Leonhard Building, University Park, PA, 16802, United States, JZJ5077@psu.edu, Soundar Kumara, Venky Shankar The main objective of this study is generating the daily delivery routes under the conditions of m depots and n vehicles in urban transportation network and some other restrictions. Considering these characteristics, we first create the cost matrix of this routing problem using google maps API to adapt the current road conditions of the operating area. To solve the problem, we generate the delivery sequence of entire network like TSP by using genetic algorithm and the we distribute the vehicle routes with the given resource and time constraints. To speed-up the TSP solution generating process, we generate and solve the multiple subnetworks problem and then link the subnetwork solutions. 362D Scheduling Contributed Session Chair: David Johnson, Purdue University, West Lafayette, IN, United States, davidjohnson@purdue.edu 1 - Using Cluster Analysis to Develop a Risk Taxonomy for Small Unmanned Aircraft System Dothang Truong, Professor, Embry-Riddle Aeronautical University, 600 S.Clyde Morris Boulevard, Daytona Beach, FL, 32114, United States, truongd@erau.edu The growing demand for Small Unmanned Aircraft System (sUAS) in the past two years has increased the number of sUAS encounters with manned aircraft or airports, raising the risk of collision. The purpose of this research is to develop a taxonomy of risk factors for sUAS operations in National Airspace System. Specifically, this research presents a process of analyzing the sUAS sighting data using cluster analysis to classify sUAS encounter incidents into different groups based on a set of selected variables in the dataset. The taxonomy will allow authorities to understand the nature, characteristics, and risk factors for each group of sUAS encounter incidents. 2 - Team Based Task Assignment for Air Cargo Operations Yong-Hong Kuo, Assistant Professor, The University of Hong Kong, Department of Industrial and Manufacturing Systems Engineering, The University of Hong Kong, Pokfulam Road, Hong Kong, yhkuo@hku.hk, Janny M. Y. Leung Our work is motivated by a real-world task assignment problem when managing air cargo operations. The tasks are assigned to teams based on considerations such as manpower requirements, team shift times, workload balance, and meal and rest break requirements. We have developed an optimization tool to aid decision- making. 3 - A Railroad Track Geometry Life Analysis Behnam Rahimikelarijani, Lamar University, Department of Industrial Engineering, Beaumont, TX, 77710, United States, brahimikelar@lamar.edu, Maryam Hamidi This paper analyses railroad track geometry degradation based on the data gathered by geometry cars. The life data is used for reliability analysis of the track. A Weibull distribution whose scale parameter is assumed to be a function of stress factors is fitted. Accumulated tonnage and average running speed are considered as the stress factors affecting degradation of the track. The analysis can next be used in minimizing life cycle cost of tracks by determining the optimal maintenance schedule. By predicting track failure and proactively maintaining them, our methodology avoids considerable costs such as revenue loss, costs of damage to other parts, and random failures. 4 - Estimating the Potential for Cost Effective Nonstructural Flood Risk Reduction in Coastal Louisiana David Richard Johnson, Assistant Professor, Purdue University, 315 N.Grant St, West Lafayette, IN, 47907-2023, United States, davidjohnson@purdue.edu, Zening Chen Louisiana’s 2017 coastal Master Plan allocates $6.1B for nonstructural flood risk reduction measures, such as elevating homes and floodproofing commercial properties. We outline the methods used to estimate the cost effectiveness of these options and present results from the Master Plan projects. We also derive the maximum potential for nonstructural risk reduction across the coast and compare optimal standards to the current standards used for flood insurance, which are based on FEMA’s estimate of the “100-year” flood depth, and to the standards used by the Master Plan. Uncertainty in future conditions is also considered. TD57

Xiangyu Cui, Shanghai University of Finance and Economics, School of Statistics and Management, Room 2218, Shanghai, 200433, China, cui.xiangyu@mail.shufe.edu.cn, Duan Li In multiperiod mean-variance framework, the investor suffers time inconsistency. The current solution schemes either assume the investor does not possess any foresight or assume the investor possesses unlimited foresight. However, in reality, the investor often has limited foresight and can only influence his own investment behaviors over a relative short time interval. Thus, we integrate the limited foresight of the investor into the model and formulate the problem as a planner-middleman-doer game. We derive the explicit expression of the equilibrium strategy, which is called limited foresight strategy, and analyze the properties of the equilibrium strategy. 4 - Stock Trading with Realization Utility in a Regime-switching Model Shengcheng Shao, The Chinese University of Hong Kong, Hong Kong, scshao@se.cuhk.edu.hk, Xuedong He We consider a stock trading model in which the stock price follows a regime- switching process with observable regimes. The agent buys and sells the stock sequentially to maximize her terminal wealth utility and realization utility, where the latter is derived from the realized gains and losses at the sale times. We solve the optimal strategy in closed form and show that the agent holds the stock at a deep loss even in the bear market but sells the stock at a gain even in the bull market. We find that the more the agent weights the realization utility, (i) the more frequently she trades and (ii) the lower the expected terminal wealth. 362C Logistics Contributed Session Chair: Jihyun Jo, Pennsylvania State University, University Park, PA, United States, JZJ5077@psu.edu 1 - Dual Channel Warehouse with Uncertain Demands Guoqing Zhang, University of Windsor, Dept of MAME, 401 Sunset Avenue, Windsor, ON, N9B 3P4, Canada, gzhang@uwindsor.ca With the emergence of online sales channels, a warehouse has come to commonly fulfil demands for both offline and online channels. To optimize operations, such a warehouse is usually divided into two areas: one for the process of fulfilling online orders and one for storing products and filling offline orders. This paper discusses multi-item product inventory policies with both stochastic online and offline demands for a dual-channel warehouse. Such a discussion has not been addressed in existing literature. Both a mathematical model and a solution approach are proposed. The impacts of uncertainty and the warehouse space have been analyzed. 2 - Robust Optimization of Biomass Logistics Supply Chain Design Babak Badri Koohi, PhD Candidate in Operations Research, Virginia Tech, Blacksburg, VA, 24061, United States, babakbk@vt.edu, Subhash C. Sarin Optimizing the design of biomass logistics supply chain plays a crucial role in utilizing biomass as a source of renewable energy. Nevertheless, there are some uncertainties associated with biomass logistics such as availability of necessary equipment, and the amount of seasonal or periodical supply and demand. In this research study, we address these uncertainties using a robust optimization approach. 3 - Identifying Noise Level in Military Logistics under Anti-access/Area-denial Environment Rosemonde Ausseil, University of New Haven, West Haven, CT, United States, rauss1@unh.newhaven.edu, Ridvan Gedik, Amy Bednar, Mark Cowan In modern warfare, the US military is increasingly faced with anti-access/area- denial challenges, which threaten its power projection operations. By analyzing the US military’s logistics, adversaries could infer and anticipate the US military’s mission objectives. Identifying which factors are predominant in determining these intentions leads us to develop a denial and deception technique of introducing noise into the US military’s movements, with the aims of obscuring its logistics activities and disrupting the adversary’s decision-making process. The mathematical model quantifies the noise and gives greater insight into what is a sufficient noise level for adversary deception. TD56

396