数据驱动跟驰模型综述

A car- following model is one of the microscopic traffic flow models that are widely focused on by transportation research and engineering. In recent years, the rapid technological advancement in information perception and acquisition, big data, and artificial intelligence, etc., has promoted the great development of data- driven car following models. Based on data science and machine learning theory, data-driven car- following models obtain the inherent law of car- following behaviors through the training, learning, iteration and evolution of real-world vehicle motion data. This paper reviews the evolution of data-driven car-following models over the past 20 years and analyzes its two research waves driven by neural network and deep learning, respectively. Three typical types of data-driven car following models and their representatives are reviewed, including traditional machine learning- based car- following models, deep learning-based car- following models, and model-data hybrid driven car-following models. Data source analysis indicates that, although a variety of high-fidelity trajectory datasets are constantly emerging, the Next Generation Simulation (NGSIM) datasets released by the United States in 2006 are still the most widely used, inparticular in recent years. Therefore, the transferability and generalization of the models are worth investigating. We also discuss from the following aspects: model input and output including how to involve more driving behavior variables, whether it is necessary to consider more behavioral variables, and whether the existing input and output can be replaced; Model testing and verification including insufficient testing, incomplete comparison, lack of unified test dataset and test standard. At last, the key factors regarding the originality and success of data- driven car- following models are discussed. It is expected that this review can help researchers better understand the past and present situations of data-driven car-following models and promote the progress of related research.