Performance Analysis of Metropolitan Bus Rapid Transit Line via Generalized Stochastic Petri Nets

Abstract

This paper addresses the problem of assessing quality measures of a public transportation line, including factors such as reliability, commercial speed, and ride comfort. Many mathematical models can simulate the behavior of a transportation system. However, these models might be overly expensive and difficult to deal with. In an attempt to overcome these limitations, a generalized stochastic Petri net (GSPN) model was developed and applied to a bus rapid transit (BRT) line in Recife, Brazil. The model yielded an explanatory power greater than 95%, and the difference between data obtained from a field survey and the results generated by the simulation was statistically insignificant. Based on this, we verified that the model is a useful tool for testing potential modifications to a system by means of simulation because it allows different operational scenarios to be quickly assessed. Moreover, the GSPN model is simple and practical to modify and implement. Furthermore, the model does not require deep knowledge of mathematical theories, making it useful for specialists of different areas, such as modelers and transport authorities, to communicate and trade information.