The launching of autonomous driving vehicle relies on enormous tests and a reasonable test evaluation system. Compared with the real vehicle test characterized by long test cycle, high cost, high risk and low coverage, virtual simulation tests are found with the advantages of infinite expansion of scenarios, arbitrary setting of traffic flow, dynamics-based parameterized vehicle simulation model and customized configuration of driver model. This paper based on international standards of L1-L2 automatic driving functions related to real world testing, delivered a methodology for test cases design. through function-based evaluation indicators and scenario-based evaluation indicators, a systematic and logical simulation test evaluation method and framework based on driving scenario is proposed.
Research on evaluation system of simulation test based on driving scenarios for L1-L2 autonomous driving
Eighth International Conference on Electromechanical Control Technology and Transportation (ICECTT 2023) ; 2023 ; Hangzhou, China
Proc. SPIE ; 12790
2023-09-07
Aufsatz (Konferenz)
Elektronische Ressource
Englisch