Agent-Based Simulation for Investigating the Safety Concerns of Electric Vehicles in the US

When operated at low speeds, electric and hybrid vehicles have created pedestrian safety concerns in congested areas of various city centers because they have relatively silent engines compared to those of internal combustion engine vehicles (ICEVs), and there is no engine noise to warn pedestrians and cyclists of an oncoming electric or hybrid vehicle. However, driver behavior characteristics have also been considered in many studies, and the high-end prices of electric vehicles (EVs) indicate that EV drivers tend to have a higher prosperity index and are more likely to receive a better education, making them more alert while driving and more likely to obey traffic rules. In this paper, the positive and negative factors associated with EV adoption and the subsequent effects on pedestrian traffic safety are investigated using an agent-based modeling approach, in which a traffic micro-simulation of a real intersection is simulated in 3D using AnyLogic software. First, the interacting agents and dynamic parameters are defined in the agent-based model. Next, a 3D intersection environment is created to integrate the agent-based model into a visual simulation, where the simulation records the number of near-crashes occurring in certain pedestrian crossings throughout the virtual time duration of a year. A sensitivity analysis is also carried out with 9,000 subsequent simulations performed in a supercomputer to account for the variation in dynamic parameters (ambient sound level, vehicle sound level, and ambient illumination). According to the analyses, EVs pose a 25% overall higher risk to pedestrian traffic safety than ICEVs do. Although this safety risk is not as high as that observed in previous statistical reports, there is still a statistically significant difference between the near-crash risks of EVs and ICEVs, which necessitates certain solutions to EVs’ silent engines. The advantage of EVs in terms of stopping sight distance is not strong enough to compensate for their high near-crash risk due to their lower degree of auditory detectability. Also, low levels of ambient illumination increase the number of pedestrians involved in near-crashes for both EVs and ICEVs.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 39p

Subject/Index Terms

Filing Info

  • Accession Number: 01680186
  • Record Type: Publication
  • Files: UTC, NTL, TRIS, ATRI, USDOT
  • Created Date: Sep 11 2018 2:47PM