We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
无人驾驶汽车换道决策虚拟仿真实验平台.
- Authors
朱靖龙; 于蕾艳; 吴宝贵; 孙德志; 胥永康; 徐 凯
- Abstract
Aiming at the problem that it is difficult for drivers to judge the appropriate time of lane change due to complex factors in the process of traditional vehicle lane change, which is easy to lead to traffic accidents, a virtual simulation experiment platform for driverless vehicle lane change decision is established. Firstly, the success/failure events of lane change in NGSIM data set are screened out, increasing the influencing factors of vehicle lane changing such as longitudinal acceleration and establishing the optimal judgment logic. Secondly, the virtual simulation experiment platform interface of driverless vehicle lane change decision based on decision algorithm tree is established by using MATLAB GUI to realize decision visualization. Finally, the vehicle lane change decision model is established through the decision tree C4.5 algorithm, and the model is trained and verified. The training results show that the longitudinal distance between the vehicle to be changed and the vehicle behind the target lane is the most sensitive factor in lane changing. The vehicle lane changing decision model based on decision tree has good feasibility, real-time and accuracy, and the accuracy rate is 97.737 6%. The virtual simulation experiment platform of driverless vehicle lane changing decision effectively improves students' exploration ability and innovation ability.
- Subjects
AUTONOMOUS vehicles; JUDGMENT (Logic); DECISION trees; TECHNOLOGICAL innovations; LANE changing; TRAFFIC accidents; VEHICLES
- Publication
Experimental Technology & Management, 2022, Vol 39, Issue 11, p99
- ISSN
1002-4956
- Publication type
Article
- DOI
10.16791/j.cnki.sjg.2022.11.017