Informs Annual Meeting 2017

MD54

INFORMS Houston – 2017

MD54

2 - Evaluation of Risks for Power Plant Operators through Reconfiguration of Price Zones in Central Western Europe (CWE) using a Recombining Scenario Tree Robin Leisen, Research Assistant, University Duisburg-Essen, Berliner Platz 6-8, Essen, 45127, Germany, robin.leisen@uni-due.de, Tim Felling, Caroline Deilen, Christoph Weber In CWE a new price zone configuration (PZC) for electricity is discussed. A frequent change of PZC affects the risk for power plant operators as their expected net present value (NPV) varies depending on the PZC. Thereby the PZC is assessed using a hierarchical cluster algorithm. Using the new zonal prices, the contribution margin is determined. The net present value results from the probability-weighted average of the contribution margin of each scenario discounted to today. Among others, the scenario tree considers time steps, grid development, fuel prices and different zonal configurations. 3 - Evaluation of Importing Countries and Transportation Mode of Energy Resources Considering Uncertainty Shigeki Toriumi, Chuo University, 1-23-27 Kasuga, Bunkyo-ku, Tokyo, 112-8551, Japan, toriumi@ise.chuo-u.ac.jp, Ryuta Takashima In this study, we evaluate country risk and transport risk of importing energy resources using two mathematical model. First, we develop a mathematical model that determines the quantity of imports considering country risk. Next, we develop a mathematical model that determines transportation mode and route considering transport risk. Then, we analyze import cost and risk using these models. As a result, for Japan, it was found that although the cost of pipeline transport is lower than that of ship transportation, the transport risk increases. 361F Arterial and Highway Traffic Operations Sponsored: TSL, Intelligent Transportation Systems (ITS) Sponsored Session Chair: Roger Lloret-Batlle, ITS Irvine, 7430 Palo Verde Road, Irvine, CA, 92617, United States, rlloretb@uci.edu 1 - Dynamic Optimal Realtime Algorithm for Isolated Signals Xiubin Wang, Associate Professor, Texas A&M.University, 3136 TAMU, College Station, TX, 77845, United States, bwang@tamu.edu, Mark Cao, Changjun Wang This work studies intersection signal control . The optimal control policy minimizes the overall intersection delay by deciding the green intervals for signal phases dynamically . It proposes an analytical model for intersection vehicle delay and derives optimal conditions for green signal switch. Two numerical algorithms are proposed: optimum based (DORAS) and queue-based heuristic respectively. Numerical tests are conducted via simulation. 2 - Connected Vehicle Enabled Proactive Signal Control for Congestion Mitigation on Arterial Corridors Hao Yang, Visiting Assistant Professor, Lamar University, 2622 Cherry Engineering Building, Beaumont, TX, 77710, United States, yhharold@gmail.com, Xing Wu, Mm Haque With increasing penetration rates of connected vehicles, signals are able to interact with real-time traffic conditions and to be operated more proactively with traffic prediction models to mitigate road congestions and to improve road capacities. This paper develops a proactive signal control system based on connected vehicles to minimize vehicle delay at multiple intersections. The system utilizes connected vehicles to predict the volumes entering the intersection accurately and to optimize phase sequence and durations of the intersections. The simulations and field experiments illustrate the effectiveness of the algorithm on reducing vehicle delays. 3 - Core-stable Queue Routing Policies for Multiserver Queue Facilities and Highway Control Roger Lloret-Batlle, PhD Candidate, ITS.Irvine, 7430 Palo Verde Road, Irvine, CA, 92617, United States, rlloretb@uci.edu, R. Jayakrishnan Transportation supply in multiserver queues and freeways is generally consumed in an FCFS basis. We explore game-theoretical solution concepts that allow agents to cooperate with each other and efficiently violate FCFS in a core-stable fashion. In particular, we develop a dynamic allocation as an n-level Stackelberg game, characterized as a partition function game which is strong-core stable. MD53

362A Challenges and Opportunities of Establishing Data-driven Agronomic Services to Smallholder Farmers Across the World Invited: Agricultural Analytics Invited Session Chair: Daniel Ricardo Jimenez, International Center for Tropical Agriculture (CIAT), KM 17 recta Cali-Palmira, KM 17 recta Cali- Palmira, Palmira, 0057, Colombia, d.jimenez@cgiar.org Co-Chair: Andrew Jarvis, a.jarvis@cgiar.org 1 - Increasing Agricultural Production and Resilience through Data Services and Decision Support Systems The farming community is currently being challenged to build resilience to climate change and to increase crop productivity through the adoption of practices and technology. Our extension efforts have showed that low adoption of decision support systems is due to the lack of engagement with farmers during the conceptualization and development process. A success farmers engagement story is the Tri-state Climate Learning Network for Row Crop Agriculture. Through bi-annual meetings, we learned that awareness on the limitations of the data, the applicability of climate information to crop management, and the co- development of usable decision support tools are key elements of farmers’ adoption. 2 - Simulation Based Fungal Disease Modeling in Agriculture using Big Data Drew Marticorena, aWhere, Inc., Broomfield, CO, United States, drewmarticorena@awhere.com The FAO estimates that over 1B MT of food are lost due to the fungal diseases yearly. In addition, associated toxins cause harmful health effects to both animals and people that consume the contaminated food. These issues are especially prevalent amongst smallholder producers. The ability to predict where and when fungal diseases are likely to occur would be of great value across the agricultural value chain and would improve the lives of smallholder producers. Using field datasets and modern parameter optimization techniques coupled with biological based models, aWhere has developed simulation based models that predict the incidence of fungal diseases with both high sensitivity and specificity. 3 - Agriculture 3.0: Crowed Sourced Data for Decision Making and Taking in Resilient Agrifood Systems Bram Govaerts, CIMMYT, Mexico City, Mexico, b.govaerts@cgiar.org Managing the hugely complex risks that are associated with the food system of the 21st century is a major challenge for decision makers in government, civil society and the private sector alike, and one that has been neglected for the past 30 years. Therefore complex agricultural innovation systems (AIS) that can support agrifood systems for nutrition, nature conservation and national and international security are required. While Knowledge Management (KM) is an important component of AIS, previous KM frameworks did not account for the fact that agricultural systems are complex systems and did not integrate innovation with KM. The results presented will show a case of an AIS that was implemented in Mexico including crowd sourcing of field scale data to develop complex decision support systems. The case presented shows that these approaches can boost performance and steer complex systems in ways that benefit all stakeholders. 4 - Data-driven Agro-climatic Services Improve Farmer Responses to Climate Variability Climate is one of the most important factors influencing the performance farming systems. In Colombia, climate variability explains between 30-60 % of rice yield. Farmers, however, make decisions in their farms including issues related to planting date, what variety to plant, or whether to plant, at best, on the basis of no information. We develop data-driven seasonal crop-climate prediction analyses that feed into the development of a climate services platform. The analyses demonstrate that the combination of skillful seasonal climate forecasts, calibrated crop models, and a forecast platform tailored to users’ needs can prove successful in establishing a climate service for agriculture. Brenda Ortiz, Associate Professor, Auburn University, 201 Funchess Hall, Auburn, AL, 36849, United States, bortiz@auburn.edu, Wendy-Lin Bartels, Clyde Fraisse, David Zierden Julian Ramirez-Villegas, University of Leeds, Leeds, United Kingdom, Julian.ramirezvillegas@gmail.com

255