Event-Triggered Control for String-Stable Vehicle Platooning

Dolk, V. S.; Ploeg, Jeroen; Heemels, W.P.M.H.

doi:10.1109/tits.2017.2738446

Cited by 187 publications

(125 citation statements)

References 52 publications

Supporting

Mentioning

125

Contrasting

Order By: Relevance

“…= E [R k |s k = s], which denotes the expected return in case policy π is followed from state s onwards. The value function v π (s) follows the Bellman equation [25], v π (s) = a π (a|s) s ,r p (s , r|s, a) [r + ζv π (s )] , (5) which can then be maximized to find the optimal state values.…”

Section: B Deep Reinforcement Learningmentioning

confidence: 99%

“…Data is exchanged in a time-triggered fashion between components of the control loops irrespective of whether an update is actually needed. In recent years, the research community in event-triggered control (ETC) has had remarkable success in showing that the amount of samples in feedback loops can be reduced significantly compared to those time-triggered approaches (see experimental studies [3]- [5] for example). In ETC, transmission of data and thus closing of the feedback loop is triggered only on certain events, e.g., an error growing too large.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Reinforcement Learning for Event-Triggered Control

Baumann

Trimpe

Zhu³

et al. 2018

2018 IEEE Conference on Decision and Control (CDC)

View full text Add to dashboard Cite

Event-triggered control (ETC) methods can achieve high-performance control with a significantly lower number of samples compared to usual, time-triggered methods. These frameworks are often based on a mathematical model of the system and specific designs of controller and event trigger. In this paper, we show how deep reinforcement learning (DRL) algorithms can be leveraged to simultaneously learn control and communication behavior from scratch, and present a DRL approach that is particularly suitable for ETC. To our knowledge, this is the first work to apply DRL to ETC. We validate the approach on multiple control tasks and compare it to model-based event-triggering frameworks. In particular, we demonstrate that it can, other than many model-based ETC designs, be straightforwardly applied to nonlinear systems. Accepted final version.

show abstract

Section: B Deep Reinforcement Learningmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep Reinforcement Learning for Event-Triggered Control

Baumann

Trimpe

Zhu³

et al. 2018

2018 IEEE Conference on Decision and Control (CDC)

View full text Add to dashboard Cite

show abstract

“…The following control design, vehicle model, and control law are taken from Dolk et al (2017). CACC systems obey two control objectives: 1) each vehicle i, of length L i , must follow the preceding vehicle i − 1 with a desired distance d r,i (t), known as individual vehicle stability, and, 2) disturbances are attenuated along the platoon, known as string stability.…”

Section: Control Designmentioning

confidence: 99%

Predictive Triggering for Distributed Control of Resource Constrained Multi-agent Systems

2019

View full text Add to dashboard Cite

A predictive triggering (PT) framework for the distributed control of resource constrained multi-agent systems is proposed. By predicting future communication demands and deriving a probabilistic priority measure, the PT framework is able to allocate limited communication resources in advance. The framework is evaluated through simulations of a cooperative adaptive cruise control system and experiments on multi-agent cart-pole systems. The results of these studies show its effectiveness over other event-triggered designs at reducing network utilization, while also improving the control error of the system.

show abstract

“…Finally, we train a Recurrent Neural Network (RNN) model that generalizes the behavioral dataset. For evaluation, we compared four different testing [9]. approaches, by generating datasets via testing, learning RNN models from the data and computing various performance measures for detecting critical behaviors in unforeseen situations.…”

Section: Introductionmentioning

confidence: 99%

“…the distance to the vehicle ahead and may use discrete, i.e. event triggered, communication to communicate desired accelerations along the platoon [9]. Besides individual vehicle stability, the most crucial goal in controller design is to guarantee so-called string stability of the platoon [27].…”

Section: Introductionmentioning

confidence: 99%

Learning a Behavior Model of Hybrid Systems Through Combining Model-Based Testing and Machine Learning

Aichernig

Bloem

Horn

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Models play an essential role in the design process of cyberphysical systems. They form the basis for simulation and analysis and help in identifying design problems as early as possible. However, the construction of models that comprise physical and digital behavior is challenging. Therefore, there is considerable interest in learning such hybrid behavior by means of machine learning which requires sufficient and representative training data covering the behavior of the physical system adequately. In this work, we exploit a combination of automata learning and model-based testing to generate sufficient training data fully automatically. Experimental results on a platooning scenario show that recurrent neural networks learned with this data achieved significantly better results compared to models learned from randomly generated data. In particular, the classification error for crash detection is reduced by a factor of five and a similar F1-score is obtained with up to three orders of magnitude fewer training samples.

show abstract

Event-Triggered Control for String-Stable Vehicle Platooning

Cited by 187 publications

References 52 publications

Deep Reinforcement Learning for Event-Triggered Control

Deep Reinforcement Learning for Event-Triggered Control

Predictive Triggering for Distributed Control of Resource Constrained Multi-agent Systems

Learning a Behavior Model of Hybrid Systems Through Combining Model-Based Testing and Machine Learning

Contact Info

Product

Resources

About