
Mechanical sciences, Год журнала: 2024, Номер 15(2), С. 613 - 631
Опубликована: Ноя. 8, 2024
Abstract. Urban traffic congestion, obstacle avoidance, and driving efficiency are the challenges faced by autonomous-vehicle path-planning technology in cities. The traditional artificial potential field (APF) algorithm is insufficient to meet requirements of safety path planning, as it easily gets trapped local optima when dealing with complex environments. Therefore, this paper proposes a novel AS-IAPF more efficiently enhance target reachability autonomous vehicles Firstly, analyzes elucidates macroscopic model, achieving effective modeling dynamic flow stability based on Lyapunov theorem classical 1D model. Thus, threshold discriminant formula for element obtained. Secondly, aforementioned formula, new proposed. mainly includes two aspects: firstly, pre-generating initial paths introducing Gaussian oscillation coefficient force fields, avoids falling into optima; secondly, using dimensional adjustment adaptively improving adjusting strength AS-APF repulsive field, further improves planning. Finally, subjected joint simulations 2D 3D scenarios different types. research results show that outperforms other algorithms same type respect comprehensive performance multiple scenario simulation experiments. In experiments three typical scenarios, proposed can drive effectively perform corresponding avoidance actions actual ahead, ultimately safe avoidance. method while considering vehicles, providing promising approach reference planning vehicles.
Язык: Английский