Modeling of Vehicle Dynamics From Real Vehicle Measurements Using a Neural Network With Two-Stage Hybrid Learning for Accurate Long-Term Prediction

Yim, Young Uk; Oh, Se‐Young

doi:10.1109/tvt.2004.830145

Cited by 36 publications

(19 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Behavior prediction of human drivers has been widely investigated, e.g. in [21], [33], [25]. As reported in the literature, human driver prediction within the ego vehicle (i.e.…”

mentioning

confidence: 99%

Model-Based Probabilistic Collision Detection in Autonomous Driving

Althoff

Stursberg

Buss

2009

IEEE Trans. Intell. Transport. Syst.

270

174

View full text Add to dashboard Cite

Abstract-Safety of planned paths of autonomous cars with respect to the movement of other traffic participants is considered. Thereto, the stochastic occupancy of the road by other vehicles is predicted. The prediction considers uncertainties originating from the measurements and the possible behaviors of other traffic participants. In addition, the interaction of traffic participants as well as the limitation of driving maneuvers due to the road geometry is considered. The result of the presented approach is the probability of a crash for a specific trajectory of the autonomous car. The presented approach is efficient as most intensive computations are performed offline, resulting in a lean online algorithm for real-time application.

show abstract

“…Behavior prediction of human drivers has been widely investigated, e.g. in [21], [33], [25]. As reported in the literature, human driver prediction within the ego vehicle (i.e.…”

mentioning

confidence: 99%

Model-Based Probabilistic Collision Detection in Autonomous Driving

Althoff

Stursberg

Buss

2009

IEEE Trans. Intell. Transport. Syst.

270

174

View full text Add to dashboard Cite

show abstract

“…Reinforcement learning methods were also introduced both utilizing off-line [26] and on-line training [27]. Different artificial neural network aided approaches were presented using nonholonomic vehicle models in the field of control design for automated parking [28], and vehicle motion prediction, where the network is trained to replicate the dynamics of a specific vehicle [29]. However, application of artificial neural networks in safety relevant systems is only possible with post filtering by traditional algorithm.…”

Section: Motion Planning For Highly Automated Road Vehicles With a Hymentioning

confidence: 99%

Motion Planning for Highly Automated Road Vehicles with a Hybrid Approach Using Nonlinear Optimization and Artificial Neural Networks

Hegedüs¹,

Bécsi²,

Aradi³

et al. 2019

SV-JME

View full text Add to dashboard Cite

Over the last decade, many different algorithms were developed for the motion planning of road vehicles due to the increasing interest in the automation of road transportation. To be able to ensure dynamical feasibility of the planned trajectories, nonholonomic dynamics of wheeled vehicles must be considered. Nonlinear optimization based trajectory planners are proven to satisfy this need, however this happens at the expense of increasing computational effort, which jeopardizes the real-time applicability of these methods. This paper presents an algorithm which offers a solution to this problematic with a hybrid approach using artificial neural networks (ANNs). First, a nonlinear optimization based trajectory planner is presented which ensures the dynamical feasibility with the model-based prediction of the vehicle's motion. Next, an artificial neural network is trained to reproduce the behavior of the optimization based planning algorithm with the method of supervised learning. The generation of training data happens off-line, which eliminates the concerns about the computational requirements of the optimization-based method. The trained neural network then replaces the original motion planner in on-line planning tasks which significantly reduces computational effort and thus run-time. Furthermore, the output of the network is supervised by the model based motion prediction layer of the original optimization-based algorithm and can thus always be trusted. Finally, the performance of the hybrid method is benchmarked with computer simulations in terms of dynamical feasibility and run-time and the results are investigated. Examinations show that the computation time can be significantly reduced while maintaining the feasibility of resulting vehicle motions.

show abstract

“…HN and ON are abbreviations for hidden neuron and output neuron, respectively. The closed loop property of such networks are important in time series prediction problems, since it enables predictions of the output an arbitrary number of steps into the future [20]. Note that u and y are not related to the control signals and outputs in other parts of this work, but merely constitutes generic variables in this example.…”

Section: A Artificial Neural Networkmentioning

confidence: 99%

Fuel-efficient Model Predictive Control for Heavy Duty Vehicle Platooning using Neural Networks

Ling

Lindsten

Ljungqvist

et al. 2018

2018 Annual American Control Conference (ACC)

View full text Add to dashboard Cite

The demand for fuel-efficient transport solutions are steadily increasing with the goal of reducing environmental impact and increasing efficiency. Heavy-Duty Vehicle (HDV) platooning is a promising concept where multiple HDVs drive together in a convoy with small intervehicular spacing. By doing this, the aerodynamic drag is reduced which in turn lowers fuel consumption. We propose a novel Model Predictive Control (MPC) framework for longitudinal control of the follower vehicle in a platoon consisting of two HDVs when no vehicleto-vehicle communication is available. In the framework, the preceding vehicle's velocity profile is predicted using artificial neural networks which uses a topographic map of the road as input and is trained offline using synthetic data. The gear shifting and mass of consumed fuel for the controlled follower vehicle is modeled and used within the MPC controller. The efficiency of the proposed framework is verified in simulation examples and is benchmarked with a currently available control solution.

show abstract

Modeling of Vehicle Dynamics From Real Vehicle Measurements Using a Neural Network With Two-Stage Hybrid Learning for Accurate Long-Term Prediction

Cited by 36 publications

References 7 publications

Model-Based Probabilistic Collision Detection in Autonomous Driving

Model-Based Probabilistic Collision Detection in Autonomous Driving

Motion Planning for Highly Automated Road Vehicles with a Hybrid Approach Using Nonlinear Optimization and Artificial Neural Networks

Fuel-efficient Model Predictive Control for Heavy Duty Vehicle Platooning using Neural Networks

Contact Info

Product

Resources

About