Kinematic and dynamic vehicle models for autonomous driving control design

By Jason Kong, Mark Pfeiffer, Georg Schildbach, Francesco Borrelli. 2015 IEEE Intelligent Vehicles Symposium (IV), 2015

789 citations recorded by the source corpus.

Subfields: Automotive Engineering, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/3c75174706613a57e26c0c50979da752ff31b564

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

In Control and Systems Engineering

Citations in the corpus

TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors (2021, methodology): “, Unscented Kalman filter [40] with kinematic bicycle model [28]).”
Computing Systems for Autonomous Driving: State of the Art and Challenges (2020, methodology): “In [152], a comparison is present to determine which of these two models is more suitable for MPC in forecast error and computational overhead.”
Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks (2018, methodology): “Traffic actors were tracked using Unscented Kalman filter (UKF) [40] with kinematic vehicle model [41], taking raw sensor data from the camera, lidar, and radar and outputting state estimates for each tracked vehicle at the rate of 10Hz.”
End to End Vehicle Lateral Control Using a Single Fisheye Camera (2018, neutral): “The simulator itself is based on the bicycle model [16], [15] which is commonly used for lateral control application due to its simplicity and its ability to represent correctly the vehicle dynamics.”
Control of Connected and Automated Vehicles: State of the Art and Future Challenges (2018, methodology): “The system dynamics f are given by the kinematic bicycle model [167, 159] after forward Euler discretization.”
Recent Advancements in End-to-End Autonomous Driving Using Deep Learning: A Survey (2023, methodology): “Modular pipelines often entail redundant computations, as each module is trained for task-specific outputs [4, 5].”
Trajectron++: Dynamically-Feasible Trajectory Forecasting with Heterogeneous Data (2020, background): “For example, one would usually model vehicles on the road using a bicycle model [26,35].”
The kinematic bicycle model: A consistent model for planning feasible trajectories for autonomous vehicles? (2017, methodology): “Reference [8] compared the predictive power of an MPC planner based on the kinematic bicycle model and a planner using a linear dynamic bicycle model, and concluded that the kinematic model performs satisfactorily.”
CoverNet: Multimodal Behavior Prediction Using Trajectory Sets (2019, methodology): “We use a standard vehicle dynamical model [24] as similar models are effective for planning at urban (nonhighway) driving speeds [23].”
Survey of Deep Reinforcement Learning for Motion Planning of Autonomous Vehicles (2020, background): “Though the kinematic model seems quite simpliﬁed, and as stated in [25], such a model can behave signiﬁcantly different from an actual vehicle, though for the many control situations, the accuracy is suitable [24].”