The real-time performance and accuracy of traffic flow prediction directly affect the efficiency of traffic flow guidance systems, and traffic flow prediction is a hotspot in the field of intelligent transportation. To further improve the accuracy of short-term traffic flow prediction, a short-term traffic flow prediction model based on traffic flow time series analysis, and an improved long short-term memory network (LSTM) is proposed. First, perform time series analysis on traffic flow data and perform smoothing and standardization processing to obtain a stable time series as model input data, which can improve the accuracy of model training and eliminate the impact of a wide range of feature values. Then, an improved LSTM model based on LSTM and bidirectional LSTM networks are established. Combining the advantages of sequential data and the long-term dependence of forwarding LSTM and reverse LSTM, the bidirectional long-term memory network (BILSTM) is integrated into the prediction model. The first layer of the LSTM network learns and predicts the input time series and further learns and trains through the bidirectional LSTM network to effectively overcome the large prediction errors. Finally, the performance of the proposed method is evaluated by comparing the predicted results with actual traffic data. The model that is proposed in this paper is compared with the long short-term memory network (LSTM) model and the bidirectional long-term memory network (BILSTM) model. The results demonstrate that the proposed method outperforms both compared methods in terms of accuracy and stability.

The boom in bike-sharing is receiving growing attention as societies become more aware of the importance of active non-motorized traffic modes. However, the low usage of this transport mode in China raises concerns. The primary objective of this study is to explore factors affecting bike-sharing usage and satisfaction degree of bike-sharing among the bike-sharing user population in China. Data were collected by a questionnaire survey in Ningbo. A bivariate ordered probit (BOP) model was developed to examine simultaneously those factors associated with both bike-sharing usage and satisfaction degree of bike-sharing among users. Marginal effects for contributory factors were calculated to quantify their impacts on the outcomes. The results showed that the BOP model can account for commonly shared unobserved characteristics within usage and satisfaction of bike-sharing. The BOP model results showed that the usage of bike-sharing was affected by gender, household bicycle/e-bike ownership, trip model, travel time, bike-sharing stations location, and users’ perception of bike-sharing. The satisfaction degree of bike-sharing was affected by household income, bike-sharing stations location, and users’ perception of bike-sharing. It is also found that bike-sharing usage and satisfaction degree are strongly correlated and positive in direction. The results can enhance our comprehension of the factors that affect usage and satisfaction degree of bike-sharing. Based on the results, some suggestions regarding planning, engineering, and public advocacy were discussed to increase the usage of bike-sharing in Ningbo, China.

In order to clearly understand the risky riding behaviors of electric bicycles (e-bikes) and analyze the riding characteristics, we review the research results of the e-bike risky riding behavior from three aspects: the characteristics and causes of e-bike accidents, the characteristics of users’ traffic behavior, and the prevention and intervention of traffic accidents. The analysis results show that the existing research methods on risky riding behavior of e-bikes mainly involve questionnaire survey methods, structural equation models, and binary probability models. The illegal occupation of motor vehicle lanes, over-speed cycling, red-light running, and illegal manned and reverse cycling are the main risky riding behaviors seen with e-bikes. Due to the difference in physiological and psychological characteristics such as gender, age, audiovisual ability, responsiveness, patience when waiting for a red light, congregation, etc., there are differences in risky cycling behaviors of different users. Accident prevention measures, such as uniform registration of licenses, the implementation of quasi-drive systems, improvements of the riding environment, enhancements of safety awareness and training, are considered effective measures for preventing e-bike accidents and protecting the traffic safety of users. Finally, in view of the shortcomings of the current research, the authors point out three research directions that can be further explored in the future. The strong association rules between risky riding behavior and traffic accidents should be explored using big data analysis. The relationships between risk awareness, risky cycling, and traffic accidents should be studied using the scales of risk perception, risk attitude, and risk tolerance. In a variety of complex mixed scenes, the risk degree, coupling characteristics, interventions, and the coupling effects of various combination intervention measures of e-bike riding behaviors should be researched using coupling theory in the future.

Road traffic accidents are one of the world’s most serious problems, as they result in numerous fatalities and injuries, as well as economic losses each year. Assessing the factors that contribute to the severity of road traffic injuries has proven to be insightful. The findings may contribute to a better understanding of and potential mitigation of the risk of serious injuries associated with crashes. While ensemble learning approaches are capable of establishing complex and non-linear relationships between input risk variables and outcomes for the purpose of injury severity prediction and classification, most of them share a critical limitation: their “black-box” nature. To develop interpretable predictive models for road traffic injury severity, this paper proposes four boosting-based ensemble learning models, namely a novel Natural Gradient Boosting, Adaptive Gradient Boosting, Categorical Gradient Boosting, and Light Gradient Boosting Machine, and uses a recently developed SHapley Additive exPlanations analysis to rank the risk variables and explain the optimal model. Among four models, LightGBM achieved the highest classification accuracy (73.63%), precision (72.61%), and recall (70.09%), F1-scores (70.81%), and AUC (0.71) when tested on 2015–2019 Pakistan’s National Highway N-5 (Peshawar to Rahim Yar Khan Section) accident data. By incorporating the SHapley Additive exPlanations approach, we were able to interpret the model’s estimation results from both global and local perspectives. Following interpretation, it was determined that the Month_of_Year, Cause_of_Accident, Driver_Age and Collision_Type all played a significant role in the estimation process. According to the analysis, young drivers and pedestrians struck by a trailer have a higher risk of suffering fatal injuries. The combination of trailers and passenger vehicles, as well as driver at-fault, hitting pedestrians and rear-end collisions, significantly increases the risk of fatal injuries. This study suggests that combining LightGBM and SHAP has the potential to develop an interpretable model for predicting road traffic injury severity.

Understanding the influence factors and related causation of hazardous materials can improve hazardous materials drivers’ safety awareness and help traffic professionals to develop effective countermeasures. This study investigates the statistical distribution characteristics, such as types of hazardous materials transportation accidents, driver properties, vehicle properties, environmental properties, road properties. In total, 343 data regarding hazardous materials accidents were collected from the chemical accident information network of China. An ordered logit regression (OLR) model is proposed to account for the unobserved heterogeneity across observations. Four independent variables, such as hazardous materials drivers’ properties, vehicle properties, environmental properties, and road properties are employed based on the OLR model, an ordered multinomial logistic regression (MLR) is estimated the OLR model parameters. Both parameter estimates and odds ratio (OR) are employed to interpret the impact of influence factors on the severity of hazardous materials accidents. The model estimation results show that 10 factors such as violations, unsafe driving behaviors, vehicle faults, and so on are closely related to accidents severity of hazardous materials transportation. Furthermore, three enforcement countermeasures are proposed to prevent accidents when transporting hazardous materials.