Quantifying and predicting driving risks for connected autonomous vehicles (CAVs) is critical to ensuring the safe operation of traffic in complex environments. This study first establishes a car-following model for CAVs based on molecular force fields. Subsequently, using a convolutional neural network and long short-term Memory (CNN-LSTM) deep-learning model, the future trajectory of the target vehicle is predicted. Risk is quantified by employing models that assess both the collision probability and collision severity, with deep-learning techniques applied for risk classification. Finally, the High-D dataset is used to predict the vehicle trajectory, from which the speed and acceleration of a target vehicle are derived to forecast driving risks. The results indicate that the CNN-LSTM model, when compared with standalone CNN and LSTM models, demonstrates a superior generalization performance, a higher sensitivity to risk changes, and an accuracy rate exceeding 86% for medium- and high-risk predictions. This improved accuracy and efficacy contribute to enhancing the overall safety of connected vehicle platoons.

To enhance road safety and optimize intelligent driving systems, this study introduces the concept of “driving style tendency” to characterize short-term driver behavior, particularly lane-changing patterns. A multidimensional framework is established to analyze driving roles and behaviors, utilizing a Hidden Semi-Markov Model and Hierarchical Dirichlet Process for the unsupervised segmentation of driving trajectory data into behavioral primitives. By systematically analyzing driver behaviors in leading and following scenarios, characteristic thresholds are derived through distribution fitting, enabling the development of a non-parametric Bayesian-based scoring method for driving style tendency. The K-means clustering algorithm is employed to transform primitive segments into quantifiable semantic information, facilitating the interpretation of driver behavior preferences. This research contributes to improved collision risk prediction in complex traffic environments, supports the design of personalized driving assistance systems, and provides valuable insights for autonomous driving technology development.