Abstract
Car-following model is indispensable to evaluate the characteristics of car-following behaviors. Through an analysis and comparison of data-driven and theoretically driven car-following models, it shows that the data-driven model has poor interpretability and high quality of training data set is required, while for the theoretical-driven model, it is unable to describe the individualized features and models of the driver so as to a low model accuracy. To solve the problem, a novel modelling method is proposed using adaptive Kalman filter algorithm to integrate the long-short-time memory neural network (LSTM) data-driven model with the IDM theoretical-driven model to build the car-following model. Test results of real driving data from a single driver prove that the fusion car-following model has higher accuracy than a single model, while at the same time highlighting the driver's personality compared to the IDM model. Besides, it improves the generalization ability of the traditional data model, which is reflected by better fitting in the extreme case (for example, the stable state when the acceleration, velocity is zero). Finally, the trajectory simulation results show that the proposed integrated data-driven car-following model can better simulate the micro-traffic behavior of car following.
Highlights
A large part of the driving process is car following, especially driving in city
long-short-time memory neural network (LSTM) uses Recurrent Neural Network (RNN) as the carrier and adds input gates, forget gates and output gates in hidden layers, which decide the filtered input information is, and the discarded information to improve the disappearance of the gradient
The basic principle is shown in Figure.4 and described as follows: Based on the theoretical framework of deep learning [19], [20], this paper adopted Python language and Keras as the development platform to construct the recurrent neural network car-following model based on LSTM unit
Summary
A large part of the driving process is car following, especially driving in city. Modelling and exploring car following behavior, quantitative analysis of the interaction between subsequent vehicles, at the macro level, helps to understand the characteristics of traffic flow, reveals the causes of traffic congestion and its temporal and spatial evolution. The theoretical driver model is dominant at the present and relatively mature, which has been well applied, for example, the vehicle safety distance model Gipps [8] used in AIMSUN and the psychophysiological Wiedemann model [9] used in VISSIM These models are mainly applied in the field of traffic flow theory and control, which represent the general driving behavior characteristics and are usually not suitable for vehicle dynamics or intelligent assisted. In the actual vehicle driving environment, the theoretical models of car-following can be calibrated based on the authentic driving data (for example, the desired speed and following distance, etc.), they can describe the general characteristics of most drivers except personality so as to an inevitable obvious error in characterizing individual drivers’ microscopic driving behavior [10].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
