Modelo de simulação híbrida para implantação de bicicleta compartilhada como alternativa ao transporte público

Abstract

Brazilian capitals are ranked among cities with the worst traffic time indices. This scenario is largely influenced by the predominant use of private vehicles in the country, leading to a less efficient public transportation system, especially for short routes. Although public transport is a viable option to relieve the excess of circulating vehicles, the travel time makes this alternative less attractive. Thus, for short-distance commutes, known as the "last mile," shared bicycles emerge as a promising alternative to reduce the travel time for users, which can be integrated into the public transport system for longer distances in major urban centers. To understand this potential, simulation proves to be a methodology capable of providing a virtual representation of the environment, allowing for modifications without the need for physical intervention in the system. In this context, the aim of the work is to develop and apply a hybrid simulation model for analyzing urban transport systems with shared bicycles. With the model, it is possible not only to measure the benefits provided by these systems but also to assist the decision-making process for the implementation of new stations and improvements in urban infrastructure. The proposed model adopts a hybrid approach, that is, combining Agent-Based Simulation (ABS) and Discrete Event Simulation (DES), focusing on the central region of Belo Horizonte, MG (Contorno Region) as the study object. For the construction phase of the computational model, the AnyLogic software was used, which made it possible to simulate nine different scenarios, considering variations in the speeds of buses and bicycles, as well as the presence or absence of bike lanes, weather conditions, and traffic levels. The model validation involved collecting field data via GPS, related to speed, location, time, and elevation of bicycle displacements in the study region. Significance tests indicated that the averages of the simulated and real data are statistically equivalent, thus validating the developed computational model. At the end of the simulation, it was possible to obtain the travel times for each type of transport. The comparative analysis between the modes revealed a significant advantage in the deployment of shared bicycles: for short distances, the reduction in travel time with the use of bicycles was up to 61% compared to buses. However, it is necessary to highlight that choosing this active mode of transportation requires physical effort, which may be a limiting factor for some users.