Trajectory planning for vehicle collision avoidance imitating driver behavior

Yang, Bin; Song, Xuewei; Gao, Zhenhai; Zhu, Naixuan

doi:10.1177/09544070211030422

Cited by 8 publications

(3 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Exponential growth and success of AI approaches bring breakthroughs by facilitating development of robots [17] and autonomous vehicles (AVs) [73]. AVs have manifold benefits such as higher traffic efficiency, a lower collision rate due to active network communication between different vehicles on the road [76] and facility to self-commute for visually and physically impaired people [49].…”

Section: Introductionmentioning

confidence: 99%

RPRP-SAP: A Robust and Precise ResNet Predictor for Steering Angle Prediction of Autonomous Vehicles

Saleem,

Asim,

Van Elst

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Steering angle controller is a core module of autonomous vehicles, where a slight miscalculation can cause severe accidents. Following safety precautions, development of robust and precise steering angle predictor is an active area of research. However, existing simulator based steering angle predictors lack in predictive performance and have not been evaluated on same benchmark datasets. Furthermore, most of them are evaluated on simulated datasets and their potential on real-world data as well as in cross-domain evaluation of both types data (real-world, simulated) remain unexplored. To accelerate and expedite research related to steering angle prediction contributions of this paper are manifold: 1) It presents two benchmark datasets that are developed using Udacity and CARLA simulators. 2) Following the need for comparative study, over both simulated datasets, it benchmarks the performance of existing predictors under 2 different evaluation settings namely: same-track and cross-track evaluation. 3) In cross-domain evaluation, it explores generalization potential of predictors by training predictors on simulated data and evaluating them on 2 realworld datasets and vice versa. 4) It presents a robust and precise steering angle predictor that utilizes skip connections for proper gradient flow among different convolutional layers. In same-track evaluation where predictors are trained and evaluated on same-track data, proposed predictor outperforms existing predictors by achieving least Mean Absolute Error (MAE) of 0.13, 0.19 and 0.065 over lake track, jungle track and CARLA based datasets, respectively. Similarly, in cross-track evaluation where predictors are trained on one track and are evaluated on other track data, once again proposed predictor outperforms existing predictors by producing average least MAE errors of 0.33 and 0.06 over Udacity and CARLA datasets, respectively. Over two real-world datasets, Sully Chen and Comma.ai, the proposed predictor demonstrates superior performance compared to existing simulator-based predictors, achieving the lowest MAE of 2.41 and 0.50, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

RPRP-SAP: A Robust and Precise ResNet Predictor for Steering Angle Prediction of Autonomous Vehicles

Saleem,

Asim,

Van Elst

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…A safe trajectory with near-optimal overall performance is planned by calculating the minimum value of the cost function. Yang et al [12] designed an evaluation function based on the proximity of the driving states of manned and unmanned vehicles. The evaluation function evaluates and filters the initial trajectory clusters to ensure optimally generated trajectories.…”

Section: Introductionmentioning

confidence: 99%

Trajectory Planning of a Semi-Trailer Train Based on Constrained Iterative LQR

Wang,

Li,

Liu

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

With the development of science and technology, self-driving technology is gradually being applied to automobile semi-trailer trains. Aiming at the problem that it is challenging to plan the traveling trajectory of an autonomous semi-trailer train, a trajectory planning algorithm based on a constrained iterative linear quadratic regulator is proposed. The constrained iterative linear quadratic regulator solves the problem that the iterative constrained linear quadratic regulator cannot deal with inequality constraints by transforming inequality constraints into interior point penalty functions to deal with all kinds of inequality constraints in the trajectory planning problem. When dealing with the obstacle avoidance problem, the body contour is modeled approximately according to the actual shape of the vehicle, and the obstacle avoidance constraints are transformed into inequality constraints. Considering the unique body structure of a semi-trailer train, the articulation angle constraint function is designed according to different vehicle speeds. Simulation experimental results show that the algorithm can plan safe driving trajectories for semi-trailer trains in complex traffic scenarios.

show abstract

“…High simultaneity performance was required for velocity planning algorithms to plan optimal trajectories quickly, in complex environmental scenarios (Chen et al, 2019;Yang et al, 2022). Continuous optimization algorithms have gained attention in recent years for their ability to define cost functions and constraint equations almost simultaneously (Consolini et al, 2017;Consolini et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

An Optimization-based Time-optimal Velocity Planning for Autonomous Driving

HU,

PAN,

GAO

et al. 2023

SIC

View full text Add to dashboard Cite

Velocity planning plays an important role in motion planning of automated driving as it must meet safety, comfort, and traffic regulation requirements. Therefore, it is necessary to consider Jerk constraint and dynamic obstacle constraint. However, the introduction of these constraints makes velocity planning a non-convex optimization problem, significantly increasing computational complexity. To address these challenges, this paper investigates an optimization-based timeoptimal velocity planning method. The non-convex and non-linear problems caused by Jerk constraint and dynamic obstacle constraint are addressed by realizing constraint linearization through velocity filtering with acceleration as the threshold. The linear programming (LP) method is then used twice to calculate a time-optimal velocity profile that satisfies the given constraints. Furthermore, when hard constraints are unable to satisfy obstacle avoidance planning, a dynamic constraint frame strategy is proposed to relax the hard constraints and fully utilize the dynamic performance of the ego-vehicle to avoid obstacles. Finally, simulations are conducted in various driving scenarios to validate the effectiveness of the proposed approach. The simulation results demonstrate that the approach proposed in this paper can quickly generate velocity profiles that meet safety and comfort constraints, within a short planning period. Additionally, the dynamic constraint frame strategy can improve the dynamic adaptability of the algorithm.

show abstract

Trajectory planning for vehicle collision avoidance imitating driver behavior

Cited by 8 publications

References 37 publications

RPRP-SAP: A Robust and Precise ResNet Predictor for Steering Angle Prediction of Autonomous Vehicles

RPRP-SAP: A Robust and Precise ResNet Predictor for Steering Angle Prediction of Autonomous Vehicles

Trajectory Planning of a Semi-Trailer Train Based on Constrained Iterative LQR

An Optimization-based Time-optimal Velocity Planning for Autonomous Driving

Contact Info

Product

Resources

About