Informs Annual Meeting 2017

TA70

INFORMS Houston – 2017

2 - Sensor Fusion and on-line Monitoring of Friction Stir Blind Riveting for Lightweight Materials Manufacturing Weihong Guo, Rutgers, The State University of New Jersey, 96 Frelinghuysen Rd, CoRE Rm 220, Piscataway, NJ, 08854, United States, wg152@rutgers.edu, Zhe Gao Friction stir blind riveting (FSBR) is a recently developed manufacturing technique for joining lightweight materials for wider adoption in the automotive sector. Using FSBR to join carbon fiber-reinforced polymer (CFRP) composite and aluminum alloy sheets has been studied experimentally, however, the quantitative relationship between FSBR and joint quality remains unclear. To gain a better understanding of FSBR lightweight materials manufacturing, the proposed method effectively models this relationship by integrating data de- noising, dimension reduction, feature extraction, feature selection, and classifier fusion. The proposed method is demonstrated with real data from FSBR. 3 - Remaining Useful Life Prediction Based on Model Integration for Broaching Tools Wenmeng Tian, Virginia Tech, 415 New Kent RD, Blacksburg, VA, 24060-6507, United States, tian0414@vt.edu, Lee J. Wells, Jaime Camelio Remaining useful life prediction is a critical task for various multi-edged machining processes such as broaching. Tool wear occurs when cutting edge and machining surface rub each other, leading to material loss from the cutting edge. A system-level health index and the parametric form of its degradation path are derived simultaneously based on the physical process model. Then the degradation path is used for remaining useful life prediction. A Bayesian update framework is used to combine in-situ observations of a testing sample and historical degradation model parameter distributions considering possible correlation between the degradation path parameters and the random failure threshold. 4 - Online Monitoring of Line Resistance in Aerosol Jet Printing Process Rao Prahalada, University of Nebraska-Lincoln, Lincoln, NE, 68588, United States, rao@unl.edu, Roozbeh Salary, Mark Poliks, Jack Lombardi The goal of this research is online monitoring of functional electrical properties, e.g., resistance, of electronic devices made using aerosol jet printing (AJP) additive manufacturing (AM) process. In pursuit of this goal, the objective is to recover the cross-sectional profile of AJP-deposited electronic traces (called lines) through shape-from-shading (SfS) analysis of their online images. The aim is to use the SfS-derived cross-sectional profiles to predict the electrical resistance of the lines. 371D OR Application in Policy Study and Environment Analysis Sponsored: Energy, Natural Res & the Environment Environment & Sustainability Sponsored Session Chair: Yihsu Chen, University of California Santa Cruz, 1156 High Street, M/S SOE3, Santa Cruz, CA, 95060, United States, yihsuchen@ucsc.edu Co-Chair: Ryuta Takashima, Tokyo University of Science, Tokyo, Japan, takashima@rs.tus.ac.jp 1 - Optimizing Water Pollution Monitoring Systems: Regulation Policy Guideline for Curbing Nutrient Pollution Michael K. Lim, Univ of Illinois Urbana-Champaign, Champaign, IL, United States, mlim@illinois.edu, Xin Chen We examine regulatory guidelines of surface water quality to curb nutrient pollution resulting from various farming activities. We formulate an optimization model that captures the government’s regulation decision taking into account farmers’ moral hazard issue. We obtain policy insights and guidelines, along with a solution method for the problem. 2 - Theoretical and Simulation Analysis of Electricity Market for Integrative Evaluation of Renewable Energy Policies from Social Welfare Aspect Masaaki Suzuki, Tokyo University of Science, 2641 Yamazaki, Noda-shi, Chiba, Japan, m-suzuki@rs.tus.ac.jp, Mari Ito, Ryuta Takashima Governments have introduced various policies for promoting renewable energy technologies such as feed-in tariff and renewable portfolio standard. Our purpose is to clarify how the relationships among policy, market power, and number of producers impact social welfare. In this work, multi-agent system is constructed for integrative evaluation of renewable energy policies. Multi-agent simulations TA69

enable us to evaluate more realistic market and to observe emergent processes of equilibrium states. By comparing the results obtained from the simulation and the equilibrium analysis, we comprehensively examine the policies from both bottom-up and top-down viewpoints. 3 - Analysis of Climate-induced Vulnerability of Northern California Natural Gas System Sepehr Ramyar, University of California-Santa Cruz, 1156 High Street, Santa Cruz, CA, 95064, United States, sramyar@ucsc.edu, Yihsu Chen, Andrew Lu Liu This talk will discuss an ongoing project that analyzes the vulnerability of the Northern California natural gas system, considering delivery contracts and engineering aspects. The study is formulated as an optimization problem that minimizes the operational and contractual energy delivery costs while at the same time accounting for the physical characteristics of the network to guarantee reliable delivery of natural gas to customers. The model will be coupled with the downscaled outputs from climate change models to evaluate its vulnerability and identify possible resilience options. 4 - When Agents Trade in Electricity Attributes, Who Benefits? Janiele Custodio, George Washington University, Washington, DC, United States, janiele@gwu.edu, Ekundayo Shittu The integration of RPS programs has been suggested as a way to accomplish renewable goals at low cost. However, issues such as market power and overlapping regulatory mechanisms are critical to the performance of such systems. The purpose of this research is to explore the viability and the implications of integrated markets for renewable certificates. We formulate a series of MPECs to analyze how imperfect competition and redundant renewable policies would affect the economics of an inter-regional market. Lastly, we estimate aggregate supply and demand curves using data from existing RPS programs to explore the viability of an integrated market for renewable certificates. 371E Data Mining Contributed Session Chair: Samuel Davis, Northeastern University, Boston, MA, United States, davis.sam@husky.neu.edu 1 - A Data Mining Approach to Identify Third Party Subrogation in the Health Insurance Industry Naveen Kumaresan, Data Mining Analyst, West Corporation, 11808 Miracle Hills Drive, Omaha, NE, 68154, United States, nkumaresan@west.com, Shruti Palasamudram Ramesh Every year, health insurance companies pay billions of dollars in excess to medical providers. An average of up to 5% of medical claims paid by insurers are the responsibility of other payers which leads to subrogation. Subrogation is the recovery of an insurance claim from a third party that was originally paid by a patient’s medical plan. Traditional subrogation process is laborious and time consuming. This paper discusses how WEST Corporation’s Center for Data Science utilized predictive analytics to predict the probability of a claim to be third party liable (TPL) with highest recovery chances. The predictive model has increased the identification of TPL claims by 5 times. 2 - Analyzing Patterns of Multiple Chronic Conditions and Their Associated Behavior in Temporal Direction using Multi-level Temporal Bayesian Network TA70

Syed Hasib Akhter Faruqui, Graduate Research Assistant, The University of Texas at San Antonio, One UTSA Circle, San Antonio, TX, 78249, United States, syed-hasib- akhter.faruqui@utsa.edu, Adel Alaeddini, Mary Jo Pugh, Jaramillo Carlos

The rising prevalence of multi-morbidity raises the complexity of health care management, Medicaid co-ordination and treatment planning. The domain knowledge of underlying pattern among the Multiple Chronic Conditions (MCC) from the existing longitudinal co-morbid patient dataset can be used for building a decision support system. Multi-Level Temporal Bayesian Networks can be used to reveal and analyze hidden patterns present in co-occurring MCC and any risk factor associated with it along the temporal abstraction. This study describes the prevalence of and patterns of co-morbidity among patients of different age groups receiving Medicaid assistance.

295