A Particle-Based Collision Probability Estimation Framework for Uncertainty-Aware Risk Evaluation in Autonomous Vehicles
Risk evaluation is a critical task for the safety of autonomous driving systems. A significant factor for evaluating risk is estimating collision probability, as it indicates the rate of exposure to collision events. To obtain an accurate representation of collision probability, possible situational...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2026 |
| País: | España |
| Institución: | Consejo Superior de Investigaciones Científicas (CSIC) |
| Repositorio: | DIGITAL.CSIC. Repositorio Institucional del CSIC |
| OAI Identifier: | oai:dnet:digitalcsic_::f5bbbb7a9dc21c73ec404dd92aff4665 |
| Acceso en línea: | http://hdl.handle.net/10261/428180 |
| Access Level: | acceso abierto |
| Palabra clave: | Safety autonomous vehicles particle filter collision probability risk |
| Sumario: | Risk evaluation is a critical task for the safety of autonomous driving systems. A significant factor for evaluating risk is estimating collision probability, as it indicates the rate of exposure to collision events. To obtain an accurate representation of collision probability, possible situational variations and uncertainty in the autonomous driving system must be considered. However, most existing approaches attempt to find simplified metrics that fail to represent uncertainties effectively. In this paper, a particle-based collision probability estimator is developed. The proposed approach employs a specialized particle system to estimate collision probability while accurately capturing uncertainties in the state space of traffic participants. Its performance, evaluated against a state-of-the-art method across multiple scenarios, demonstrates its ability to effectively capture diverse collision events and represent the underlying uncertainty distribution. |
|---|