汽车工程学报2024，Vol.14Issue(6)：959-969,11.DOI:10.3969/j.issn.2095‒1469.2024.06.04
基于微观车流的拟人化换道决策规划
A Humanized Lane-Changing Decision-Making and Planning Method Based on a Microscopic Traffic Flow Model
摘要
Abstract
To improve the safety and comfort of autonomous vehicles during lane changes,the proposed approach incorporates the impact of lane-changing on local traffic flow and introduces a lane-changing inertia factor based on traditional decision-making models.To overcome the limitations of decoupled longitudinal and lateral trajectory planning,a joint constraint planning approach is proposed.Using dynamic programming and quadratic programming algorithms,the current lateral trajectory curvature is adjusted based on the longitudinal constraints from the previous frame.In the longitudinal planning process,key obstacles are filtered based on the current lateral planning results,and curvature-based speed constraints are applied.The results from real vehicle experiments show that the proposed approach generates lateral and longitudinal lane-changing trajectories that effectively enhance the rationality,stability,and comfort of lane changes.关键词
智能驾驶/微观车流/换道行为决策/横纵向联合约束规划Key words
intelligent driving/the microscopic traffic car/lane-changing behavior decision-making/horizontal and vertical joint constrained programming分类
交通工程引用本文复制引用
付翔,裴超,万佳琦,江学良,王文举..基于微观车流的拟人化换道决策规划[J].汽车工程学报,2024,14(6):959-969,11.基金项目
国家重点研发计划项目(2022YFC3006005):高性能水陆两栖艇救援关键技术研究 （2022YFC3006005）
