Aggregating Data for the Flow-Intercepting Location Model: A Geographic Information System, Optimization, and Heuristic Framework: Aggregating Data for Flow-Intercepting Problems

Flow-intercepting problems have received considerable interest, represented by about 40 academic publications, since the early 1990s. Point-based demand aggregation also has received much research interest in both industry and academia. Systematic studies of flow data aggregation for flow-intercepting problems have not, however, been reported to date. Our research highlights the importance of flow-based demand aggregation and develops a framework for aggregating such demand. This framework represents the first systematic study of aggregation for flow-intercepting location models (FILM). The standard FILM is the perfect model for our goals—its aggregation errors are easy to understand and its outputs are easy to measure and compare. Our research uses geographic information systems, optimization, and heuristic technologies to examine the special network flow structure of a real-world transportation system and to develop a comprehensive method of aggregating data for the standard FILM. We apply our method to the 2001 afternoon peak traffic data for Edmonton, Alberta (the sixth largest Canadian city) and find this application to be extremely efficient. We discover that in the Edmonton traffic flow network, a large number of paths have very small flows; major flows are concentrated in a limited number of paths; and a large number of small-flow paths and a large number of low-flow nodes on local streets have negligible effects on facility locations for FILM. We speculate that most real-world transportation systems may have similar characteristics. Los problemas de tipo Flow-intercepting (intercepcion de flujo) han recibido el interes considerable de la comunidad academica desde principios de los anos 90. Muestra de ello son las casi 40 publicaciones realizadas durante dicho periodo. Los metodos de agregacion de la demanda de tipo puntual (point-based demand aggregation) tambien han sido objeto de investigacion tanto en medios industriales como academicos. En contraste, hasta la fecha, no hay registro de estudios sistematicos sobre la agregacion de datos de flujos para problemas tipo flow-intercepting. El articulo presente pone de relieve la importancia de la agregacion de la demanda basada en flujos y desarrolla un marco operativo para la agregacion de dicha demanda. Este marco constituye el primer estudio sistematico de agregacion dentro de los modelos de localizacion tipo Flow-intercepting. El modelo de localizacion Flow-intercepting estandar (FILM) es el modelo perfecto para las metas de este estudio –sus errores de agregacion son sencillos de entender y sus resultados son facilmente medibles y comparables. La investigacion utiliza Sistemas de Informacion Geograficos (SIGs), metodos de optimizacion, y procedimientos heuristicos para examinar la estructura de flujo de una red de un sistema de transporte en el mundo real y para desarrollar un metodo integral que agregue los datos para un FILM estandar. El metodo propuesto fue aplicado a datos de trafico automotor a la hora pico de la tarde del ano 2001 en la ciudad de Edmonton, Alberta (la sexta mas grande de Canada) .Los resultados obtenidos demuestran que la aplicacion del metodo es extremadamente eficiente. Asimismo este estudio muestra que en la red de trafico de Edmonton, hay un gran numero de rutas con flujos de magnitudes muy pequenas, y que los principales flujos se concentran en un numero limitado de rutas. Asimismo, se hallo que un gran numero de rutas con flujos pequenos y un gran numero de nodos con flujos pequenos correspondientes a calles locales no tienen sino efectos insignificantes sobre la ubicacion de las instalaciones para el modelo FILM. Los autores especulan que la mayoria de sistemas de transporte del mundo tienen potencialmente caracteristicas similares a las halladas en el ejemplo presentado en este estudio. 自从20世纪90年代早期以来,截流选址问题就备受关注,已发表了约40篇与此相关的学术成果。工业界和学术界也对基于点的集计需求表现出浓厚的兴趣。但迄今尚未有关于截流选址问题中流数据集计方法系统研究的报道。本研究突出了基于流的集计需求的重要性,并开发了集成此类需求的框架。此框架首次给出了截流选址模型集成的系统研究。标准截流选址模型(FILM)因其集成误差易于理解、输出易于度量和比较,因而是满足我们研究目标的理想模型。本研究采用了地理信息系统(GISs)、优化和探索性分析技术来检验现实世界中运输系统的专用网络流结构,并提出了用于标准FILM数据集计的综合方法。将此方法应用于阿尔伯达省埃德蒙顿市(加拿大第六大城市)2001年全年下午高峰运输数据,结果表明该方法极为有效。研究发现,埃德蒙顿市交通流网络中多数路径的流量非常小,而主要流量集中于少数路径;就FILM而言,大量的小流量路径和当地街道的低流量节点对设施选址影响甚微。据此我们推断,多数现实世界的运输系统具有类似特征。

[1]  Oded Berman,et al.  Locating Discretionary Service Facilities Based on Probabilistic Customer Flows , 1995, Transp. Sci..

[2]  Francisco G. Benitez,et al.  An approach to estimating and updating origin-destination matrices based upon traffic counts preserving the prior structure of a survey matrix , 2005 .

[3]  C. Revelle,et al.  Heuristic concentration: Two stage solution construction , 1997 .

[4]  M. J. Hodgson A Flow-Capturing Location-Allocation Model , 2010 .

[5]  Mark W. Horner,et al.  Network flow-based strategies for identifying rail park-and-ride facility locations , 2007 .

[6]  Michael Kuby,et al.  Prospects for geographical research on alternative-fuel vehicles , 2006 .

[7]  M. John Hodgson,et al.  Locating Vehicle Inspection Stations to Protect a Transportation Network , 2010 .

[8]  Alan T. Murray,et al.  The Influence of Data Aggregation on the Stability of p-Median Location Model Solutions , 2010 .

[9]  Qian Wang,et al.  Algorithms for a Facility Location Problem with Stochastic Customer Demand and Immobile Servers , 2002, Ann. Oper. Res..

[10]  Oded Berman,et al.  Locating Discretionary Service Facilities, II: Maximizing Market Size, Minimizing Inconvenience , 1995, Oper. Res..

[11]  M. John Hodgson,et al.  AGGREGATION ERROR EFFECTS ON THE DISCRETE-SPACE p-MEDIAN MODEL: THE CASE OF EDMONTON, CANADA , 1997 .

[12]  Timothy J. Lowe,et al.  A Synthesis of Aggregation Methods for Multifacility Location Problems: Strategies for Containing Error , 1999 .

[13]  Justin C. Williams,et al.  Solving the maximal covering location problem with heuristic concentration , 2008, Comput. Oper. Res..

[14]  Timothy J. Lowe,et al.  Demand Point Aggregation for Location Models , 2002 .

[15]  Oded Berman,et al.  Locating Multiple Competitive Facilities: Spatial Interaction Models with Variable Expenditures , 2002, Ann. Oper. Res..

[16]  Robert B. Dial,et al.  A PROBABILISTIC MULTIPATH TRAFFIC ASSIGNMENT MODEL WHICH OBVIATES PATH ENUMERATION. IN: THE AUTOMOBILE , 1971 .

[17]  John N. Hooker,et al.  Finite Dominating Sets for Network Location Problems , 1991, Oper. Res..

[18]  E. Hillsman,et al.  Errors in measuring distances from populations to service centers , 1978 .

[19]  J. Current,et al.  Analysis of Errors Due to Demand Data Aggregation in the Set Covering and Maximal Covering Location Problems , 2010 .

[20]  Gilbert Laporte,et al.  Heuristics for the location of inspection stations on a network , 2000 .

[21]  M. John Hodgson,et al.  A network location-allocation model trading off flow capturing andp-median objectives , 1992, Ann. Oper. Res..

[22]  Timothy J. Lowe,et al.  Aggregation error for location models: survey and analysis , 2009, Ann. Oper. Res..

[23]  Obed Berman The maximizing market size discretionary facility location problem with congestion , 1995 .

[24]  Michael Kuby,et al.  Optimization of hydrogen stations in Florida using the Flow-Refueling Location Model , 2009 .

[25]  M. John Hodgson,et al.  Applying the flow-capturing location-allocation model to an authentic network: Edmonton, Canada , 1996 .

[26]  James R. Evans,et al.  Aggregation and Disaggregation Techniques and Methodology in Optimization , 1991, Oper. Res..

[27]  Oded Berman,et al.  Flow-Interception Problems , 1995 .

[28]  Oded Berman,et al.  Locating flow-intercepting facilities: New approaches and results , 1995, Ann. Oper. Res..

[29]  Oded Berman,et al.  Flow intercepting spatial interaction model: a new approach to optimal location of competitive facilities , 1998 .

[30]  M. John Hodgson,et al.  The Pickup Problem: Consumers' Locational Preferences in Flow Interception , 2009 .

[31]  Polly Bart,et al.  Heuristic Methods for Estimating the Generalized Vertex Median of a Weighted Graph , 1968, Oper. Res..

[32]  M. John Hodgson Developments in Flow-Based Location-Allocation Models , 1998 .

[33]  Michael Kuby,et al.  Location of Alternative-Fuel Stations Using the Flow-Refueling Location Model and Dispersion of Candidate Sites on Arcs , 2007 .

[34]  S. L. Hakimi,et al.  Optimum Locations of Switching Centers and the Absolute Centers and Medians of a Graph , 1964 .

[35]  I. Averbakh,et al.  Locating flow-capturing units on a network with multi-counting and diminishing returns to scale , 1996 .

[36]  Oded Berman,et al.  Optimal Location of Discretionary Service Facilities , 1992, Transp. Sci..

[37]  M. Kuby,et al.  A Model for Location of Capacitated Alternative-Fuel Stations , 2009 .

[38]  D. Krass,et al.  Generalized flow-interception facility location models with probabilistic customer flows , 1997 .

[39]  M. John Hodgson,et al.  Testing a Bicriterion Location-Allocation Model with Real-World Network Traffic: The Case of Edmonton, Canada , 1997 .

[40]  O. Berman Deterministic flow-demand location problems , 1997 .

[41]  Peter A. Rogerson,et al.  Location coverage models with demand originating from nodes and paths: Application to cellular network design , 2008, European Journal of Operational Research.

[42]  Charles ReVelle,et al.  Central Facilities Location , 2010 .

[43]  M. John Hodgson,et al.  Heuristic concentration for the p-median: an example demonstrating how and why it works , 2002, Comput. Oper. Res..

[44]  Michael Kuby,et al.  The flow-refueling location problem for alternative-fuel vehicles , 2005 .

[45]  Tai-Hsi Wu,et al.  Solving the competitive discretionary service facility location problem , 2003, Eur. J. Oper. Res..