This paper compared the expected accuracy of the gray GM (1, 1) model and the combined GMBP model using a data set for major road traffic accidents. A combined GMBP prediction model composed of the very first parameter gray model GM (1, 1) is able to make exact predictions for forecasting dreary type of processes, and BP (back-propagation) neural network for a major traffic accident is proposed to overcome the limitations of a single prediction model for a major traffic accident. The method first obtains predicted data using the gray GM (1, 1) model, then trains the BP neural network using the GM (1, 1) model’s predicted data and the original data as input and output data, respectively, and finally the trained BP neural network can be considered a combined GMBP prediction model. The predicted data of the digit of major traffic accidents, the digit of fatalities, and the digit of damages from 2008 to 2020 were obtained using the combined GMBP prediction model, and it is compared with the expected data of the single gray GM (1, 1) model. The grades showed that the exactitude of the combined GMBP prediction model was significantly higher than that of the single gray GM (1, 1) model. Finally, from 2021 to 2033, the combined GMBP prediction model is used to forecast the number of significant road traffic incidents, mortalities, and damages. The prediction results show that the number of accidents, faculties, and injuries is on the decline in the future years.

Take-over time is a critical factor affecting safety. Accurately predicting the take-over time provides a more reliable basis on issuing take-over requests, assessment of take-over risks, and optimization of human–machine interaction modes. Although there has been substantial research on predicting take-over time, there are still shortcomings in personalized prediction (particularly in accounting for individual differences in driving experience, cognitive abilities, and physiological responses). To gain a comprehensive understanding of the characteristics and applicability of take-over time prediction methods, this review covers four aspects: literature search information, factors influencing take-over time, data acquisition and processing methods, and take-over time prediction methods. Through literature search, research hotspots in recent years have been summarized, revealing the main research directions and trends. Key factors influencing take-over time, including driver factors, autonomous driving systems, and driving environments, are discussed. Data preprocessing stages, including data acquisition and processing, are systematically analyzed. The advantages and disadvantages of classical statistical, machine learning, and cognitive architecture models are summarized, and the shortcomings in current research are highlighted (for instance, the limited generalizability of models trained predominantly on simulator data to real-world driving scenarios). By thoroughly summarizing the strengths and weaknesses of existing research, this review explores under-researched areas and future trends, aiming to provide a solid theoretical foundation and innovative research perspectives for optimizing take-over time prediction, thereby promoting the widespread application and efficient development of autonomous driving technology.

Existing studies had shown that advanced driver assistance systems (ADAS) and driver individual characteristics can significantly affect driving behavior. Therefore, it is necessary to consider these factors when building the car‐following model. In this study, we established a car‐following model based on risk homeostasis theory, which uses safety margin (SM) as the risk level quantization parameter. Firstly, three‐way Analysis of Variance (ANOVA) was used to analyze the influencing factors of car‐following behavior. The results showed that ADAS and driving experience have a significant effect on the drivers’ car‐following behavior. Then, according to these two significant factors, the car‐following model was established. The statistical method was used to calibrate the parameter reaction response τ. Other four parameters (SM_DL, SM_DH, α₁, and α₂) were calibrated using a classical genetic algorithm, and the effects of ADAS and driving experience in these four parameters were analyzed using T‐test. Finally, the proposed model was compared with the GHR model, and the result showed that the proposed model has a smaller Root Mean Square Error (RMSE) than the GHR model. The proposed model is a method of simulating different driving behaviors that are affected by ADAS and individual characteristics. Considering more driver individual characteristics, such as driving style, is the future research goal.