Publication
Trajectory planning based on adaptive model predictive control: Study of the performance of an autonomous vehicle in critical highway scenarios
| Summary: | Increasing automation in automotive industry is an important contribution to overcome many of the major societal challenges. However, testing and validating a highly autonomous vehicle is one of the biggest obstacles to the deployment of such vehicles, since they rely on data-driven and real-time sensors, actuators, complex algorithms, machine learning systems, and powerful processors to execute software, and they must be proven to be reliable and safe. For this reason, the verification, validation and testing (VVT) of autonomous vehicles is gaining interest and attention among the scientific community and there has been a number of significant efforts in this field. VVT helps developers and testers to determine any hidden faults, increasing systems confidence in safety, security, functional analysis, and in the ability to integrate autonomous prototypes into existing road networks. Other stakeholders like higher-management, public authorities and the public are also crucial to complete the VTT process. As autonomous vehicles require hundreds of millions of kilometers of testing driven on public roads before vehicle certification, simulations are playing a key role as they allow the simulation tools to virtually test millions of real-life scenarios, increasing safety and reducing costs, time and the need for physical road tests. In this study, a literature review is conducted to classify approaches for the VVT and an existing simulation tool is used to implement an autonomous driving system. The system will be characterized from the point of view of its performance in some critical highway scenarios. |
|---|---|
| Main Authors: | Gonçalves, Catarina Isabel Vaz |
| Subject: | Autonomous Vehicles Safety-Critical Simulation Testing Validation Verification |
| Year: | 2022 |
| Country: | Portugal |
| Document type: | master thesis |
| Access type: | open access |
| Associated institution: | Universidade da Beira Interior |
| Language: | English |
| Origin: | uBibliorum |
| Summary: | Increasing automation in automotive industry is an important contribution to overcome many of the major societal challenges. However, testing and validating a highly autonomous vehicle is one of the biggest obstacles to the deployment of such vehicles, since they rely on data-driven and real-time sensors, actuators, complex algorithms, machine learning systems, and powerful processors to execute software, and they must be proven to be reliable and safe. For this reason, the verification, validation and testing (VVT) of autonomous vehicles is gaining interest and attention among the scientific community and there has been a number of significant efforts in this field. VVT helps developers and testers to determine any hidden faults, increasing systems confidence in safety, security, functional analysis, and in the ability to integrate autonomous prototypes into existing road networks. Other stakeholders like higher-management, public authorities and the public are also crucial to complete the VTT process. As autonomous vehicles require hundreds of millions of kilometers of testing driven on public roads before vehicle certification, simulations are playing a key role as they allow the simulation tools to virtually test millions of real-life scenarios, increasing safety and reducing costs, time and the need for physical road tests. In this study, a literature review is conducted to classify approaches for the VVT and an existing simulation tool is used to implement an autonomous driving system. The system will be characterized from the point of view of its performance in some critical highway scenarios. |
|---|
Funded activities
No funded projects found.