Integrating spatial autocorrelation into location-allocation problems

点击翻译按钮获取中文摘要

This project is about improving the speed and quality of solutions for network facility location problems, such as locating new fire stations to reduce fire loss or closing existing schools in population-declining communities. These challenges called location-allocation (L-A) problems are extremely difficult to solve optimally. This research investigates the tendency for similar or dissimilar attribute values, such as residential housing concentrations to cluster geographically near facility service usage. This clustering is called spatial autocorrelation (SA). These researchers will fill a significant gap in the spatial analysis literature by determining how SA information can facilitate more robust solutions in L-A analyses. This new approach will enable better and faster computational solutions to large L-A problems for location both public and private facilities that are translatable to other research domains, providing economical benefits to society. The literature describing relationships between SA and locational solutions to L-A problems is scant. This research seeks to analyze a strongly suspected relationship between SA in facility demand variables and solutions to L-A problems. It will exploit this relationship to reformulate L-A models and their computer solution methods. Revised problem formulations will focus on the reduction of the number of possible locational solutions examined so that solving large-size L-A problems and other related locational problems becomes tractable and efficient. This investigation of the novel interface between spatial statistics (e.g., SA) and spatial optimization (e.g., L-A) will provide new knowledge and convincing evidence that should spur further research in various disciplines, including operations research, regional science, and geography.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.