Long memory is important: A test study on deep-learning based car-following model

Wang, Xiao; Jiang, Rui; Li, Li; Lin, Yilun; Wang, Fei‐Yue

doi:10.1016/j.physa.2018.09.136

Cited by 78 publications

(47 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Asymmetric CF behavior characteristics were effectively captured based on LSTM which outputs velocity and distance gap, and driving memory was emphasized and incorporated in different time scales [15]. Wang et al [16] applied GRU neural networks to study the relationship between long memory effect and hysteresis phenomena in congested freeway traffic, inferring that long memory is important and should be embedded in CF models. The importance of historical driving memory in CF modeling has also been acknowledged in some other studies [3].…”

Section: Cf Models Considering Historical Driving Memorymentioning

confidence: 99%

“…. For prediction pattern 2, its corresponding models are the main data-driven and deep learning models such as RNN, LSTM, and GRU [3,15,16]. Unlike the majority of existing CF models which take the instantaneous velocity, relative velocity, and distance headway as inputs, these advanced models use all information observed in the last few seconds and the output variables are not fixed.…”

Section: Model Setups For Cf Modelingmentioning

confidence: 99%

See 1 more Smart Citation

Car-Following Modeling Incorporating Driving Memory Based on Autoencoder and Long Short-Term Memory Neural Networks

et al. 2019

View full text Add to dashboard Cite

Although a lot of work has been conducted on car-following modeling, model calibration and validation are still a great challenge, especially in the era of autonomous driving. Most challengingly, besides the immediate benefit incurred with a car-following action, a smart vehicle needs to learn to evaluate the long-term benefits and become foresighted in conducting car-following behaviors. Driving memory, which plays a significant role in car-following, has seldom been considered in current models. This paper focuses on the impact of driving memory on car-following behavior, particularly, historical driving memory represents certain types of driving regimes and drivers’ maneuver in coordination with the variety of driving regimes. An autoencoder was used to extract the main features underlying the time-series data in historical driving memory. Long short-term memory (LSTM) neural network model has been employed to investigate the relationship between driving memory and car-following behavior. The results show that velocity, relative velocity, instant perception time (IPT), and time gap are the most relevant parameters, while distance gap is insignificant. Furthermore, we compared the accuracy and robustness of three patterns including various driving memory information and span levels. This study contributes to bridging the gap between historical driving memory and car-following behavior modeling. The developed LSTM methodology has the potential to provide personalized warnings of dangerous car-following distance over the next second.

show abstract

Section: Cf Models Considering Historical Driving Memorymentioning

confidence: 99%

Section: Model Setups For Cf Modelingmentioning

confidence: 99%

Car-Following Modeling Incorporating Driving Memory Based on Autoencoder and Long Short-Term Memory Neural Networks

et al. 2019

View full text Add to dashboard Cite

show abstract

“…There is a commonly accepted view that considering historical information is crucial for behavior recognition and accurate trajectory prediction [1][2][3][4]. To support this view, many researchers have proven that an instantaneous driving operation can generate a long-term impact on vehicle behavior (movement), and this phenomenon is called the memory effect [1][2][3]. The memory effect is an implicit mechanism during driving, which cannot be directly observed.…”

Section: Introductionmentioning

confidence: 99%

“…for trajectory prediction. These models are easy to calculate, but they do not consider the uncertainties of the observed states [8] and cannot distinguish driving fluctuations 1) from vehicle behavior changes due to the lack of historical information, resulting in unreliable long-term predictions.…”

Section: Introductionmentioning

confidence: 99%

Investigating the dynamic memory effect of human drivers via ON-LSTM

Dai

Cao

et al. 2020

Sci. China Inf. Sci.

Self Cite

View full text Add to dashboard Cite

Investigating temporal trends in the explanatory variables related to the severity of drivers' injuries in single-vehicle collisions Journal of Traffic and Transportation Engineering (English Edition) 4, 71 (2017); Dynamic and steady control of quantum coherence in photonic crystals via the Zeeman effect SCIENCE CHINA Physics, Mechanics & Astronomy 55, 2300 (2012); A two-dimensional simulation method for investigating charge transport behavior in 3-D charge trapping memory

show abstract

“…From the perspective of theoretical research, the situation as lots of traffic congestion is imbalanced for the traffic system. To explain the evolution traffic congestion formation and improve the stability of traffic flow, scholars at home and abroad have proposed a series of traffic models in recent decades, such as the car-following models [1,2], lattice hydrodynamic models [3,4], and cellular automaton models [5][6][7][8][9].…”

Section: Introductionmentioning

confidence: 99%

A New Two-Lane Lattice Hydrodynamic Model considering the Traffic Interruption Probability under Honk Environment

Cheng

2020

Complexity

View full text Add to dashboard Cite

By accounting for the traffic interruption probability on two-lane highway under honk environment, an extended lattice hydrodynamic model is presented in the paper. In view of the novel model, a series of researches are carried out. The neutral stability condition and the stable condition can be derived through linear analysis. Then, the mKdV equation near the critical point can be obtained by applying nonlinear analysis, which describes the traffic jams according to the kink-antikink density waves. In addition, numerical simulation is performed, which confirms that traffic interruption probability on two-lane highway under honk environment can develop traffic jams by the change of density waves. Also, the phenomenon is consistent with the results of previous theoretical analysis. It shows that accounting for the traffic interruption probability on two-lane highway under honk environment can stabilize the traffic flow efficiently.

show abstract

Long memory is important: A test study on deep-learning based car-following model

Cited by 78 publications

References 44 publications

Car-Following Modeling Incorporating Driving Memory Based on Autoencoder and Long Short-Term Memory Neural Networks

Car-Following Modeling Incorporating Driving Memory Based on Autoencoder and Long Short-Term Memory Neural Networks

Investigating the dynamic memory effect of human drivers via ON-LSTM

A New Two-Lane Lattice Hydrodynamic Model considering the Traffic Interruption Probability under Honk Environment

Contact Info

Product

Resources

About