Computational Framework for Verifiable Decisions of Self-Driving Vehicles
Computational Framework for Verifiable Decisions of Self-Driving Vehicles
复制标题
自动驾驶车辆可验证决策的计算框架
DOI:
10.1109/ccta.2018.8511432
复制
发表时间:
2018
期刊:
影响因子:
--
通讯作者:
S. Veres
中科院分区:
文献类型:
--
作者:
Mohammed Al;Hongyang Qu;S. Veres
A framework is presented for the verification of an agent's decision making in autonomous driving applications by checking the logic of the agent for instability and inconsistency. The framework verifies the decisions of a rational agent implemented in Natural Language Programming (NLP) and based on a belief-desire-intention (BDI) paradigm using sEnglish and Jason code. The main results are methods of verification for the correctness of real-time agent decisions expressed in computational tree logic (CTL) formulae. The methods rely on the Model Checker for Multi-Agent Systems (MCMAS) verification tool. To test the new verification system, an autonomous vehicle (AV) has been modelled and simulated, which is capable of planning, navigation, objects detection and obstacle avoidance using a rational agent. The agent's decisions are based on information received from mono-cameras and LiDAR sensor that feed into logic-based decisions of the AV. The model of the AV and its environment has been implemented in the Robot Operating System (ROS) and the Gazebo virtual reality simulator.
登录
查看更多内容
影响因子:
3.3
作者:
N. Lincoln;S. Veres
通讯作者:
N. Lincoln;S. Veres
影响因子:
9
作者:
Lincoln N
通讯作者:
Lincoln N
DOI:
10.48550/arxiv.1611.03372
发表时间:
2016
期刊:
arXiv e-prints
影响因子:
--
作者:
Izzo Paolo
通讯作者:
Izzo Paolo
DOI:
10.1109/ecc.2014.6862528
发表时间:
2014
期刊:
--
影响因子:
--
作者:
Qu H
通讯作者:
Qu H