Real-Time Heuristic Motion Planning for Autonomous Vehicle Driving

貢獻的翻譯標題: 即時啟發式運動路徑規劃於車輛自動駕駛應用

Shyr Long Jeng*, Wei Hua Chieng, Yung Chen Wang


研究成果: Article同行評審

2 引文 斯高帕斯(Scopus)


This study presents a novel method that can be used to perform dynamic motion planning for a nonholonomic four-wheeler vehicle according to actual dynamic constraints to achieve a rapid response in the driving environment and obtain optimal vehicle performance. The proposed intuitive steering model employs inferential heuristics of driving behavior and relevant two-direction indices obtained from a driver's-perspective view instead of traditional path planning to quickly and efficiently generate target paths without the requirement of complicated mathematical calculations. In the full model of a nonholonomic vehicle, a chassis with Ackerman steering geometry and double-wishbone suspension, lateral and longitudinal forces from contact strains, and forces due to the suspension system are considered to verify the motion of an autonomous vehicle from its starting position to its destination. Simulations of lane changes and U-turn maneuvers reveal that the proposed method provides a feasible trajectory in real time. The path planning method is effective for many driving scenarios, provides high dynamic performance, and maintains high maneuverability. This method is intuitive and extensible, thus allowing vehicles to navigate in real time and adapt to dynamic environmental conditions.

頁(從 - 到)187-196
期刊Journal of the Chinese Society of Mechanical Engineers, Transactions of the Chinese Institute of Engineers, Series C/Chung-Kuo Chi Hsueh Kung Ch'eng Hsuebo Pao
出版狀態Published - 4月 2021