Automated Modeling and Simulation of Nonhomogeneous Poisson Processes Using SAS and Visual Basic

• Activity Number: 42
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Hualiang Zhao*+ and James R. Wilson and Michael E. Kuhl
• Companies: North Carolina State University and North Carolina State University and Rochester Institute of Technology
• Keywords: estimation ; simulation ; nonhomogeneous Poisson processes ; multiresolution procedure ; SAS ; Visual Basic

Abstract

In practical applications of industrial and systems engineering techniques, we often encounter complex arrival processes in which the arrival rate depends on time. Examples include arrivals of patients at an emergency room, visits to a Web site, and arrivals of phone calls at a customer service center. To understand and improve system operation in those situations, nonhomogeneous Poisson processes (NHPPs) have often been used to model such time-varying arrival processes. We formulate and implement a flexible multiresolution procedure for estimating the mean-value function of a given NHPP based on one or more independent realizations of the process. An efficient algorithm for simulating the fitted NHPP is also developed using the SAS software package. The estimation and simulation procedures are integrated into a software system with a user-friendly graphical interface built with Visual Basic.