Real-time Model Based Path Planning for Wheeled Vehicles

Jordan, Julian; Zell, Andreas

doi:10.1109/icra.2019.8794133

Cited by 4 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the literature for vehicle motion planning has many variations [7], [20]- [25], most of the algorithms emulate a hierarchical architecture of behavior planning and trajectory generation. To have a valid baseline that represents the level of performance in the literature, we used state-of-the-art methods for each layer, such as Intelligent Driving Model (IDM) [26] and Minimizing-Overall-Braking Induced-by-Lane-changes (MOBIL) [27] algorithms for behavior planner (BP) and a global optimization for trajectory generation.…”

Section: Planning On Frenet Spacementioning

confidence: 99%

An End-to-end Deep Reinforcement Learning Approach for the Long-term Short-term Planning on the Frenet Space

Moghadam,

Alizadeh,

Tekin

et al. 2020

Preprint

View full text Add to dashboard Cite

Tactical decision making and strategic motion planning for autonomous highway driving are challenging due to the complication of predicting other road users' behaviors, diversity of environments, and complexity of the traffic interactions. This paper presents a novel end-to-end continuous deep reinforcement learning approach towards autonomous cars' decision-making and motion planning. For the first time, we define both states and action spaces on the Frenet space to make the driving behavior less variant to the road curvatures than the surrounding actors' dynamics and traffic interactions. The agent receives time-series data of past trajectories of the surrounding vehicles and applies convolutional neural networks along the time channels to extract features in the backbone. The algorithm generates continuous spatiotemporal trajectories on the Frenet frame for the feedback controller to track. Extensive high-fidelity highway simulations on CARLA show the superiority of the presented approach compared with commonly used baselines and discrete reinforcement learning on various traffic scenarios. Furthermore, the proposed method's advantage is confirmed with a more comprehensive performance evaluation against 1000 randomly generated test scenarios.

show abstract

Section: Planning On Frenet Spacementioning

confidence: 99%

An End-to-end Deep Reinforcement Learning Approach for the Long-term Short-term Planning on the Frenet Space

Moghadam,

Alizadeh,

Tekin

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…In [25], it was postulated that Hybrid A* could be extended to consider cost based on a very simple traversability model, but that approach has not been evaluated on the basis of either simulations or real experiments. Recently, Hybrid A* was used in terrain driving for a small UGV with a high-fidelity vehicle model [26] by applying a receding horizon strategy. However, this method has a short lookahead distance and neglects to provide a cost-to-go estimate when prioritizing nodes, which limits the planning horizon.…”

Section: Introductionmentioning

confidence: 99%

“…While the use of a traversability grid search as a basis for a sampling-based search that considers kinematic constraints has been attempted before, the previous approaches either lack a sufficiently long planning horizon [26], [23] or do not consider optimality in the sampling search [28], [27]. To address these problems, in this paper, we develop the Traversability Hybrid A* algorithm.…”

Section: Introductionmentioning

confidence: 99%

Path Planning for UGVs Based on Traversability Hybrid A*

Thoresen

Nielsen

Mathiassen

et al. 2021

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

In this paper, a new method of path planning for unmanned ground vehicles (UGVs) on terrain is developed. For UGVs moving on terrain, path traversability and collision avoidance are important factors. If traversability is not considered, the planned path may lead a UGV into areas that will cause rough vehicle motion or lead to the UGV getting stuck if the traversability is low. The proposed path planning method is based on the Hybrid A* algorithm and uses estimated terrain traversability to find the path that optimizes both traversability and distance for the UGV.The path planning method is demonstrated using simulated traversability maps and is compared to the original Hybrid A* algorithm. The method is also verified through real-time experiments in real terrain, further demonstrating the benefits of terrain traversability optimization using the proposed path planning method. In the experiments, the proposed method was successfully applied for autonomous driving over distances of up to 270 m in rough terrain. Compared with the existing Hybrid A* method, the proposed method produces more traversable paths.

show abstract