Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MA34

With Mobility-on-Demand system serves as the first-and-last-mile solution, the public transit can greatly extend its coverage, thereby alleviating congestion on the roads. In the context of such a hybrid system, due to the imbalanced nature between customer pick-up and drop-off, rebalancing the vehicles is necessary to ensure adequate performance. This study proposes an approximate dynamical programming solution to optimally making the vehicle relocation decisions, aiming at minimizing the number of rebalanced vehicles. Extensive simulation results show that the proposed strategy can significantly reduce the customer walk-aways and allow the system to operate with a smaller fleet size. 2 - Evaluating Spatial Pricing in Ride Sourcing Systems a Graph Fused Lasso Denoising Approach Natalia Zuniga-Garcia, Graduate Research Assistance, University of Texas at Austin, Department of Civil, Architectural and Environmental Engineering, 301 E. Dean Keeton St. Stop C1761, Austin, TX 78712, United States., Austin, TX, 78712, United States, Mauricio Garcia-Tec, James G. Scott, Natalia Ruiz-Juri, Randy B. Machemehl This study explores the spatial pricing discrimination of ride-sourcing trips using empirical data from over 1 million rides in Austin, Texas. We implement the graph-fussed lasso (GFL) technique as a total variation denoising method. GFL globally smooths anisotropic discrete data while maintaining local adaptivity. GFL smoothing is posed as a convex optimization problem, and we solve it using a novel flexible and scalable algorithm with high computational efficiency. Our results include a temporal and spatial exploration of different ride-sourcing operational and productivity variables in Austin. 3 - Dynamic Traffic Assignment of Autonomous Mobility On-demand Rongsheng Chen, University of Minnesota, Minneapolis, MN, 55455, United States, Michael W. Levin This study explores the traffic conditions of a city road network with many travelers served by autonomous mobility-on-demand. The focus is to model the effect of rebalancing traffic which is caused by the unbalanced distribution of travelers and vehicles. A modified dynamic traffic assignment which includes car- sharing trip chains is used to find user equilibrium behavior. A linear program with flow conservation constraints is used to calculate the number of rebalancing trips with minimum total travel cost and to generate the dispatch strategy. Optimization Direct Vendor Demo Session 1 - Analyzing Unstructured Text Data with the JMP Pro 14 Text Explorer Mia L. Stephens, SAS Institute Inc, P.O. Box 290, York Harbor, ME, 03911, United States In the era of big data, a majority of the data captured by organizations is unstructured. Much of this unstructured data is in the form of text - from customer feedback, survey results, emails and texts, web reports, social media and other channels. Analyzing this text-based information is particularly challenging, but the new Text Explorer platform in JMP 14 makes it easy. This platform provides an efficient and interactive tool for analyzing unstructured text data, allowing us to easily extract information and transform unstructured text data into structured information. In this session, we’ll use case studies to demonstrate how to use the JMP Text Explorer platform to analyze unstructured text data. We’ll use a word cloud to visualize word frequency, use latent class analysis to cluster words, and apply other tools to understand underlying themes in unstructured text data. We’ll also see how to create a document term matrix (DTM), and will use the resulting structured data in predictive modeling. 2 - A DOCplex and ODH|CPLEX Python Primer Robert Ashford, Optimization Direct, Harrington Park, NJ, United States This short tutorial shows participants how to build a basic model using the DOCplex API in python. This session includes setting the python environment, reading data from a csv or spreadsheet, creating variables, objective functions, constraints, solving the model, and returning the results. Additionally this session points the participants to further reading so that they may expand their capabilities. Furthermore we will present the brand new ODH|CPLEX API for python, which improves solution times for large models. n MA34 North Bldg 223 8:00 - 8:45 JMP. A division of SAS /8:45 - 9:30

n MA32 North Bldg 222B Alternative-Fuel Vehicles for Sustainable Transportation Sponsored: TSL/Freight Transportation & Logistics Sponsored Session Chair: Mesut Yavuz, University of Alabama, Tusc, AL, 35487, United States Co-Chair: Isil Koyuncu, The University of Alabama, Tuscaloosa, AL, 35406, United States 1 - Optimizing Placement of Electric Vehicle Charging Stations and Incentives to Combat Overstaying Ragavendran Gopalakrishnan, Postdoctoral Associate, Cornell University, Ithaca, NY, United States, Arpita Biswas, Partha Dutta In this two-part talk, we first discuss optimizing placement of Electric Vehicle (EV) charging stations with both budget and coverage constraints, by introducing an iterative heuristic to solve a mixed packing-and-covering problem by alternately invoking knapsack and set-cover algorithms. Next, with increasing adoption of EVs, overstaying after active charging lowers utilization, and warrants imposing penalties. Higher penalties could be unacceptable to risk-averse users due to uncertain parking durations, leading to decreased utilization (and revenue). Our solution, validated on London charging data, increases both utilization and revenue while significantly reducing overstaying. 2 - Charging Station Network Design for Electrified Vehicles in Urban Communities Seyed Sajjad Fazeli, Wayne State University, Detroit, MI, United States, Saravanan Venkatachalam, Ratna Babu Chinnam, Alper E. Murat The societal benefits of large-scale adoption of EVs cannot be realized without the adequate deployment of accessible charging stations due to mutual dependence of EV sales and public infrastructure deployment. Such infrastructure deployment also presents some unique opportunities for promoting livability within communities. In this research, we develop models and methods to improve community’s access to the EV charging stations. A choice modeling approach embedded into two-stage stochastic programming model is proposed to determine the optimal network of charging stations including type, capacity, and location of electric charging stations based on EV drivers preference. 3 - Financial Analysis of Electric Vehicle Fast Charging: A Case Study in San Diego California Eleftheria Kontou, National Renewable Energy Laboratory, 15013 Denver West Parkway, Golden, CO, 80401, United States The objective of this study is financial analysis of electric vehicle (EV) fast charging stations. We focus on a case study in San Diego CA and we utilize three models to simulate chargers utilization over the planning horizon of 2018 to 2025, analyze operational costs of electricity under an EV pricing scheme, and estimate profitability indices and net present values for several scenarios of 125 and 400kW plugs stations. We explore the impact of the station’s number of plugs, power level and power-sharing, energy storage sizes, and photovoltaic array collocation on the financial viability of such EV fast charging stations providers. 4 - Green Inventory Routing Problem for Perishable Products Gokce Palak, Shenandoah University, 1460 University Dr., Winchester, VA, 22601, United States We consider a multi-period multi-vehicle green inventory routing problem for perishable products. Through a numeric study, we analyze the impacts of age- dependent perishability, carbon emissions, and a heterogeneous fleet on inventory management and routing decisions. n MA33 North Bldg 222C Autonomous Vehicles: Service Design, Management, and Control Sponsored: TSL/Intelligent Transportation Systems (ITS) Sponsored Session Chair: Rongsheng Chen, University of Minnesota, Minneapolis, MN, United States 1 - An Optimization Approach for Vehicle Re-balancingin a Mobility- on-demand System Liu Xu, University of Maryland-College Park, 1173 Glenn L. Martin Hall, College Park, MD, 20742, United States, Xinlei Zhang, Ali Haghani

131