Proactive Traffic Management Research Articles

An advanced real-time crash prediction framework for combined hard shoulder running and variable speed limits system using transformer

One of the important applications of real-time crash prediction analysis lies in the field of proactive traffic management, where instantaneous crash risk evaluation and dynamic decision-making are prerequisites. This research proposes an integrated and advanced real-time crash risk prediction framework for Variable Speed Limits (VSL) and Hard Shoulder Running (HSR) implemented freeways considering their operational periods. Statistical methods are utilized to identify the significant crash contributing factors (related to traffic, roadway geometry, and weather conditions) and explain their relationships with crashes. Time-Embedded Transformer models are proposed to predict the likelihood of real-time crash events. The sensitivity and false alarm rate of the proposed AM peak model are found to be 0.76 and 0.27, respectively, whereas the values are 0.78 and 0.24, respectively for the PM peak model. Additionally, the results indicate substantial improvements in model prediction performance (i.e., an increment of sensitivity values by 7.04% and 8.33% in the AM and PM models, respectively) after incorporating the general safety condition of a freeway segment as an input feature while estimating real-time crash prediction models. Practitioners and policymakers could apply this method to obtain a more accurate real-time crash likelihood estimation, identify important crash precursors, dynamically update algorithms, and enhance safety aspects while operating traffic management strategies on freeways.

A conflict risk graph approach to modeling spatio-temporal dynamics of intersection safety

Intersections are among the most hazardous roadway spaces due to the complex and conflicting road users’ movements. Spatio-temporal modeling of conflict risks among road users can help to identify strategies to mitigate the exacerbation of safety risks and restore hazardous conditions to normal traffic situations. This paper proposes the 'Conflict Risk Graph' as a novel concept to infer real-time conflict risks at intersections at a fine-grained level by mapping conflict-prone locations to nodes within a network characterized by specific topological structures. A significant contribution of this work is the development of a Transformer-based Graph Convolutional Network (Trans-GCN), a model that synergistically combines the Transformer's proficiency in capturing global dependence with the GCN's ability to learn local correlations. The Trans-GCN adeptly models the complex evolution patterns of conflict risks at intersections. The evaluation in this paper against five common state-of-the-art deep learning approaches demonstrates the superior performance of the Trans-GCN in conflict risk inference and adaptability to node changes. Furthermore, extensive experiments with different node configurations reveal a correlation between node setup and model performance, showing that higher spatio-temporal resolution decreases inference accuracy. This insight informs the selection of an optimal node configuration that balances the detailed capture of spatio-temporal dynamics with modeling accuracy, enabling ideal conflict risk inferences at intersections. Ultimately, this work offers significant insights for the enhancement of proactive traffic safety management and the advancement of intelligent traffic systems.

AI-Enabled Traffic Light Control System: An Efficient Model to Manage the Traffic at Intersections using Computer Vision

Traffic congestion is a significant issue with studies indicating it costs cities billions annually and averages 54 hours of wasted time per traveler each year. This situation necessitates the implementation of efficient traffic management systems, especially at intersections. In response to this challenge, our work introduces an artificial intelligence-based system designed to analyze and predict traffic flow using machine learning algorithms and deep learning methods in conjunction with traffic cameras. The model comprises two main components: real-time data collection and predictive modeling. It employs object detection to identify and classify vehicles and adjusts traffic signal timings based on the necessary passage time and predetermined constraints. Additionally, data accumulated during operation facilitates the development of a predictive model for traffic flow over time, allowing for proactive traffic management. Evaluations are done to showcase the accuracy of the model and corresponding simulation and physical implementation further approved the applicability of our approach. Finally, this work aims to enhance urban transportation efficiently, reduce commuting stress, and improve the quality of life for city residents

Graph convolutional LSTM algorithm for real-time crash prediction on mountainous freeways

Accurate real-time traffic crash prediction is crucial for proactive traffic safety management. Currently, the majority of real-time models forecast crashes every 5 min to support different intelligent transportation systems. However, these intervals might be too short for practical use in manually implementing proactive traffic safety measures such as deploying traffic law enforcement and emergency rescue resources. Therefore, this study develops hourly crash prediction models to provide network operators with sufficient time to take measures in advance. A section of a mountainous freeway in Guizhou province is divided into homogeneous segments, with crash data, traffic operations data, and meteorological data being collected hourly. As the result is an imbalanced dataset of crash and non-crash instances, the training dataset is resampled using synthetic minority over-sampling technique (SMOTE) to address the issue. To fully capture the complex spatiotemporal relationships in the data and achieve high crash prediction accuracy, a graph convolutional network-long short-term memory (GCN-LSTM) model is constructed for the first time, combining a graph convolutional network (GCN) and long short-term memory (LSTM) neural network. For comparison purposes, LSTM, Extreme Gradient Boosting (XGBoost), and logistic regression (LR) models are developed. The results show that the GCN-LSTM model outperforms other models in hourly traffic crash prediction, and the optimal prediction performance is achieved with the crash-to-non-crash ratio of 1:4. The GCN-LSTM method is found to effectively capture the complex spatiotemporal relationships in prediction data and to handle imbalanced traffic crash data.

A data mining-then-predict method for proactive maritime traffic management by machine learning

Proactive traffic management is increasingly critical in maritime intelligent transportation systems. Central to this is maritime traffic forecasting, which leverages specific structures and properties of the problem. This study focuses on the traffic dynamics within convergent areas of inland waterways and proposes a method based on data mining followed by prediction using Automatic Identification System (AIS) data. This approach addresses uncertainties in ship voyage destinations and optimizes predictions for temporary stops in inland waterways. AIS data is processed to depict complete ship motion trajectories, grouping them into trajectory sets based on shared origin, destination, and route. These groups help represent maritime traffic patterns using the entrance and exit points of channels and the boundaries of the study area. Additionally, a stop detection model is applied to these trajectories to identify nodes within maritime traffic networks. A decision tree algorithm is then employed to train a classifier for predicting traffic patterns. The method was validated in the convergent area of the Yangtze River and the Hanjiang River, demonstrating effective pattern extraction from inland maritime traffic and high accuracy in predicting single ship trajectories, achieving a 96.7% accuracy rate and 80.9% precision. The findings suggest that the proposed method (1) effectively extracts and predicts traffic patterns, (2) identifies congestion in convergent waters, and (3) supports traffic management strategies.

A crash occurrence risk prediction model based on variational autoencoder and generative adversarial network

Traffic crash risk prediction is pivotal for proactive traffic safety management but faces challenges due to the extreme imbalance between crash and non-crash data. This paper proposes integrating variational autoencoder into a generative adversarial network (VAE-GAN) for crash data augmentation without information loss. VAE-GAN generates higher-quality data due to its superior deep generative capabilities. For the crash occurrence risk prediction task, we utilize a convolutional neural network (CNN) trained on the balanced datasets generated by VAE-GAN. Two kinds of category determinations are tested for better prediction results by using prediction probability maximum and selecting threshold. The prediction model integrating variational autoencoder into a deep convolutional generative adversarial network (VAE-DCGAN) exhibits the best performance. Furthermore, we leverage Shapley additive explanations (SHAP) to interpret the key features and patterns impacting prediction results. This analysis gains insights into the underlying mechanisms of crash occurrences, and helps improve proactive traffic management and control strategies.

A Review on Highway Network System - Samruddhi Expressway

Highway networks serve as critical arteries of transportation infrastructure, facilitating the movement of goods and people across vast geographical regions. As traffic volumes continue to increase and technological advancements reshape transportation systems, there is a pressing need to optimize the efficiency, safety, and sustainability of highway networks. This abstract presents a comprehensive system approach aimed at addressing these challenges. The proposed highway network system integrates advanced technologies such as intelligent transportation systems (ITS), real-time traffic monitoring, data analytics, and machine learning algorithms to optimize traffic flow, minimize congestion, and enhance safety. By leveraging data from various sources including sensors, cameras, and connected vehicles, the system enables proactive traffic management and congestion prediction, allowing authorities to implement timely interventions and mitigate potential disruptions. Furthermore, the integration of artificial intelligence and machine learning algorithms enables predictive maintenance of infrastructure, identifying potential issues before they escalate into critical failures. This proactive approach not only reduces maintenance costs but also enhances the reliability and lifespan of highway assets. Moreover, the system prioritizes safety through the implementation of automated collision avoidance systems, adaptive speed control, and real-time incident management. By leveraging vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication technologies, the system enables vehicles to exchange critical information, such as hazardous road conditions or impending collisions, thereby reducing the risk of accidents and improving overall safety. Additionally, the system incorporates sustainability principles by optimizing traffic flow patterns to minimize fuel consumption and emissions, promoting eco-friendly transportation practices, and facilitating the integration of electric and autonomous vehicles into the existing infrastructure.

Predictive Analysis of Maritime Congestion Using Dynamic Big Data and Multiscale Feature Analysis

The maritime industry is one of the most crucial sectors in the global economy, facilitating the transportation of goods and commodities across vast distances. However, maritime network congestion has become an increasingly critical challenge that significantly affects shipping efficiency and the overall operational performance of the industry. This study proposes an innovative congestion prediction approach using dynamic big data analysis of vessel trajectories and multiscale feature analysis. First, the dynamic analysis of vessel trajectories aims to extract valuable information from ships’ data as they navigate the oceans, enabling proactive traffic management and optimized routing. Second, the multiscale feature analysis provides a comprehensive understanding of maritime network congestion by examining it from different perspectives and scales, leading to more accurate predictions and effective congestion management strategies. Furthermore, this study introduces an enhanced Faster R‐CNN vessel detection model for real‐time tracking, integrating convolutional and SKNet networks. To improve short‐term traffic flow prediction accuracy, this study employs multiscale feature analysis through wavelet transformation. The foundational traffic data undergo wavelet decomposition for a detailed representation across frequencies. Gated recurrent unit (GRU) neural network and autoregressive moving average (ARMA) models are utilized to predict trend and noise components, respectively. Fusion of predictions demonstrates superior accuracy and is validated against real data. Empirical results showcase minimal errors and heightened prediction accuracy compared to actual data.

Prediction of Traffic Congestion in Vehicular Ad-Hoc Networks Employing Extreme Deep Learning Machines (Edrlm)

Vehicular Ad-Hoc Networks (VANETs) represent a crucial component of intelligent transportation systems (ITS), enabling vehicles to communicate with each other and with roadside infrastructure. Predicting traffic congestion in VANETs is essential for enhancing road safety, optimizing traffic flow, and improving overall transportation efficiency. Traditional machine learning methods have shown promise in this domain; however, they often fall short in handling the complex, high-dimensional data typical of VANETs. To address these challenges, this study employs Extreme Deep Learning Machines (EDRLM), an advanced deep learning technique, for traffic congestion prediction. The EDRLM framework leverages the strengths of deep neural networks and extreme learning machines, offering a robust and scalable solution for processing the dynamic and heterogeneous data in VANETs. By integrating feature extraction, selection, and prediction into a unified model, EDRLM can capture intricate patterns and temporal dependencies within traffic data. The proposed model is trained and validated using real-world VANET datasets, incorporating various traffic parameters such as vehicle speed, density, and inter-vehicular distances. Our experimental results demonstrate that EDRLM outperforms conventional machine learning algorithms in terms of prediction accuracy, computational efficiency, and robustness to noise and missing data. The model's ability to provide timely and precise congestion predictions can facilitate proactive traffic management strategies, including dynamic routing and adaptive traffic signal control, ultimately leading to reduced travel times and enhanced road safety. This study underscores the potential of EDRLM in transforming traffic management in VANETs, paving the way for more intelligent and adaptive ITS solutions. Future research directions include exploring hybrid models combining EDRLM with other advanced machine learning techniques and expanding the framework to accommodate emerging vehicular communication technologies such as 5G and Internet of Things (IoT) devices.

Dynamic Bayesian hierarchical peak over threshold modeling for real-time crash-risk estimation from conflict extremes

Using traffic conflict-based extreme value theory (EVT) models to quantify real-time crash-risk of road facilities is a promising direction for developing proactive traffic safety management strategies. Existing EVT real-time crash-risk analysis studies have only focused on using block maxima models. This study proposes a dynamic Bayesian hierarchical peak over threshold modeling approach to estimate real-time crash-risk based on traffic conflicts. The proposed approach combines quantile regression, dynamic updating approach, Bayesian hierarchical structure, and the peak over threshold method to generate time-varying generalized Pareto distributions to derive real-time crash-risk measures (i.e., crash probability and return level). The derived real-time crash-risk measures are applied to estimate cycle-level crash-risk at three signalized intersections in Surrey, British Columbia. Five approaches are used to dynamically update the model parameters, including time trend model, generalized autoregressive conditional heteroskedasticity process approach, as well as the first-order, second-order, and third-order dynamic linear models. For comparison, static models are also developed. All the developed models are compared in terms of statistical fit and predictive performance. Based on the best fitted dynamic model, cycle-level crash probability and return level are calculated to measure signalized intersection safety at cycle level. The results show that dynamic models considerably outperform static models in terms of statistical fit and predictive performance. Further, the third-order dynamic model has the best performance, which is probably due to that the model incorporates two linear trends to respectively describe the variation of the coefficients as well as its change to better account for the variation in the effect of time-varying covariates. However, it should be noted that the third-order dynamic model development needs more computation time than other dynamic models, which may limit the application of the model.

Real-Time Big Data Analytics and Proactive Traffic Safety Management Visualization System

Big data and data-driven analysis could be utilized for traffic management to improve road safety and the performance of transportation systems. This paper introduces a web-based proactive traffic safety management (PATM) and real-time big data visualization tool, which is based on an award-winning system that won the US Department of Transportation (USDOT) Solving for Safety Visualization Challenge and was selected as one of the USDOT Safety Data Initiative (SDI) Beta Tools. State-of-the-art research, especially for real-time crash prediction and PATM, are deployed in this study. A significant amount of real-time data is accessed by the system in order to conduct data-driven analysis, such as traffic data, weather data, and video data from closed-circuit television (CCTV) live streams. Based on the data, multiple modules have been developed, including real-time crash/secondary crash prediction, CCTV-based expedited detection, PATM recommendation, data sharing, and report generation. Both real-time data and the system outputs are visualized at the front end using interactive maps and various types of figures to represent the data distribution and efficiently reveal hidden patterns. Evaluation of the real-time crash prediction outputs is conducted based on one-month real-world crash data and the prediction results from the system. The comparison results indicate excellent prediction performance. When considering spatial-temporal tolerance, the sensitivity and false alarm rate of the prediction results [i.e., high crash potential event (HCPE)] are 0.802 and 0.252, respectively. Current and potential implementation are also discussed in this paper.

An ensemble learning method for low visibility prediction on freeway using meteorological data

AbstractThe prediction of low visibility is essential for proactive traffic safety management on freeways under fog conditions. However, few studies have developed prediction models for visibility on freeways at a short‐term time interval. This study proposes an ensemble learning approach to develop a short‐term prediction model of low visibility on freeways using meteorological data. Spearman's rank correlation coefficient is used to select meteorological variables related to low visibility. Random forests (RF) and extreme gradient boosting (XGB) are employed to develop visibility prediction models, and back propagation neural network (BPNN) and logistic regression (LR) are used for comparison. The models are evaluated over five prediction time intervals (5, 10, 15, 30, and 60 min). The results indicate that the RF models outperform the other models with precision of 73.9%, recall of 59.8% and F1 score of 0.65. Moreover, the prediction model with a 15‐min time interval shows better performance. With the proposed short‐term prediction of low visibility, it is expected that more crashes could be prevented with more appropriate proactive traffic safety management strategies.

New Modeling Approach for Predicting Disaggregated Time-Series Traffic Crashes

Short-window crash prediction is a fundamental step in proactive traffic safety management that can monitor traffic conditions in real time, identify unsafe traffic dynamics, and implement suitable interventions for traffic conflicts. Short-window (e.g., hourly) traffic-collision count data, however, exhibits excessive zeros and serial autocorrelation. Most of the commonly used regression-based models fail to address excessive zeros and temporal structure simultaneously in hourly traffic-collision prediction. For example, Hurdle models and zero-inflated (ZI) models can address the overdispersion issue caused by excessive zeros but lack power to control the significant spatiotemporal characteristics inherent in the time-based collision data. To overcome these issues simultaneously, this paper develops a novel statistical model termed Zero-Inflated Logarithmic link for count Time series (ZILT) which is based on the framework of ZI models. Covariates (e.g., speed, vehicle type, and traffic volume) were extracted through deep-learning computer vision methods in vehicle detection and tracking on the image space. This new statistical model (i.e., ZILT) performs better at solving the issues of excessive zeros and serial dependencies. The prediction accuracy of the ZILT model improved by around 5% in relation to zero-inflated Poisson (ZIP) and Hurdle models. Results show that traffic crashes happening in the previous hour and other covariates such as truck-to-car ratio, holiday effect, traffic flow, and speed have significant influence on collision occurrence. Findings from this study could be utilized by relevant transport agencies in developing engineering interventions and countermeasures to proactively manage road safety.

A methodology for prioritizing safety indicators using individual vehicle trajectory data

A methodology for assessing crash risk using vehicle driving trajectories based on data mining techniques was developed in this study. A variety of safety indicators reflecting the characteristics of traffic and road geometric conditions were evaluated in terms of their capability of capturing hazardous traffic flow. Comprehensive data preparation was conducted by matching driving trajectory data obtained from in-vehicle digital tachograph devices and crash data to classify and analyze hazardous and normal traffic flows. The random forest approach was adopted to quantify the importance of safety indicators. The crash risks were evaluated using the logistic regression model and multivariate adaptive regression splines model based on the set of safety indicators with high importance. The results show that the dangerous driving events rate and driving volatility indicators were found to be particularly significant in identifying hazardous conditions. The multivariate adaptive regression splines model showed better performance and a classification accuracy of 86% was achieved. The proposed methodology will be useful for deriving effective countermeasures to prevent crashes, which is the backbone of proactive traffic safety management.

A Traffic Flow Dependency and Dynamics based Deep Learning Aided Approach for Network-Wide Traffic Speed Propagation Prediction

The information of network-wide future traffic speed distribution and its propagation is beneficial to develop proactive traffic congestion management strategies. However, predicting network-wide traffic speed propagation is non-trivial. This study develops a traffic flow dependency and dynamics based deep learning aided approach (TD2-DL), which predict network-wide high resolution traffic speed propagation by explicitly integrating temporal-spatial flow dependency, traffic flow dynamics with deep learning method techniques. Specifically, we first develop a graph theory-based method to identify the local temporal-spatial traffic dependency of each road among neighboring roads adaptive to the prediction horizon and traffic delay. Then, traffic speed propagation on every road is mathematically described by v-CTM based on traffic initial and boundary conditions. Next, the long short-term memory (LSTM) model is employed to predict boundary conditions factoring the traffic temporal-spatial dependency and historical data predicted by v-CTM. In this way, we well couple the physical models (traffic dependency and v-CTM) with the deep learning approach, and further make them coevolution under this framework. Last, an EKF is used to assimilate predicted traffic speed predicted by v-CTM coupled with the LSTMs and the field traffic data; an FNN is introduced to impute missing and corrupted data for improving the traffic speed prediction accuracy. The numerical experiments indicated that the TD2-DL predicted the network-wide traffic speed propagation in 30 minutes with accuracy varying from 85%-98%. It outperformed the tested models recently developed in literature. The ablation experimental results confirmed the significance of factoring traffic dependency and integrating data imputation and assimilation techniques for improving the prediction accuracy.

Feature engineering methodology for congestion forecasting

Short-term traffic forecasting is a key element in proactive traffic management, e.g., mitigating the negative effect of impending congestion through appropriate capacity allocation at signalized intersections. In this study, we develop a data-driven methodology for reliably and robustly predicting impending stable congestion. By incorporating feature engineering techniques into an iterative machine learning process, we develop a prediction model that can be intuitively understood by traffic experts and is amenable to diagnostics during implementation. Our iterative machine learning process combines the embedded and filter approaches for feature selection with the use of expert knowledge to create aggregative input variables. The embedded approach is represented by application of a decision tree algorithm, while the filter approach is reflected in use of the mean decrease in accuracy output of a random forest algorithm for identifying expressive variables. We tested the methodology by applying it to field data from a sub-network in Tel Aviv. We demonstrated a reduction in the number of decision tree input variables from 66 raw variables to the five most effective aggregative ones, while achieving statistically significant improvement in all performance indicators. The identification rate of stable congestion increased from 65% to 74% while the robustness of the results was enhanced: the standard deviations of the identification and false alarm rates fell from 8% to 3%, respectively, to 5% and 2%. • Short-term traffic forecasting is a key element in mitigating the negative effect of impending urban congestion. • A machine learning process for reliably predicting impending congestion was developed by using feature engineering. • Aggregative input variables were created by combining embedded and filter approaches with the use of expert knowledge. • The proposed methodology significantly increases both the congestion identification rate and the robustness of the results. • The final prediction model can be intuitively understood by traffic experts.

Wasserstein Generative Adversarial Network to Address the Imbalanced Data Problem in Real-Time Crash Risk Prediction

Real-time crash risk prediction models aim to identify pre-crash conditions as part of active traffic safety management. However, traditional models which were mainly developed through matched case-control sampling have been criticised due to their biased estimations. In this study, the state-of-art class balancing method known as the Wasserstein Generative Adversarial Network (WGAN) was introduced to address the class imbalance problem in the model development. An extremely imbalanced dataset consisted of 257 crashes and over 10 million non-crash cases from M1 Motorway in United Kingdom for 2017 was then utilized to evaluate the proposed method. The real-time crash prediction model was developed by employing Deep Neural Network (DNN) and Logistic Regression (LR). Crash predictions were performed under different crash to non-crash ratios where synthetic crashes were generated by Wasserstein Generative Adversarial Network (WGAN), Synthetic Minority Over-sampling Technique (SMOTE) and Adaptive Synthetic (ADASYN) sampling respectively. Comparisons were then made with algorithmic-level class balancing methods such as cost-sensitive learning and ensemble methods. Our findings suggest that WGAN clearly outperforms other oversampling methods in terms of handling the extremely imbalanced sample and the DNN model subsequently produces a crash prediction sensitivity of about 70% with a 5% false alarm rate. Based on the findings of this study, proactive traffic management strategies including Variable Speed Limit (VSL) and Dynamic Messing Signs (DMS) could be deployed to reduce the probability of crash occurrence.

Hidden Mixture Vehicle Discharge State Inference at Signalized Intersection Using Vehicle Travel Time and Discharge Headway Data

Accurate and reliable traffic state identification is crucial to developing responsive and proactive traffic management applications. In this study, the problem of vehicle discharge state identification at signalized intersections is investigated, which focuses on the vehicle discharge process during the green interval. Instead of using detailed vehicle trajectory data and treating the observations of vehicles independently, this study formulates the vehicle discharge process in a Hidden Markov Model (HMM) framework using sequential observations of vehicle travel time and discharge headway as inputs. Three vehicle discharge states (i.e., overflow, single stop, and free arrival) are encoded as latent states, and a restricted left-to-right state transition matrix is imposed to respect the nature of the vehicle discharge process in the real world. The standard HMM is further extended to incorporate two informative covariates to parameterize the probabilities of the initial states and state transitions. The proposed models have been validated on the Next Generation Simulation (NGSIM) dataset. Compared to a benchmark model, the proposed models show their strength in correctly inferring the vehicle discharge state and are more reliable to use in presence of random missing observations. The effectiveness of covariate incorporation is also investigated, and several extended applications of the proposed models are discussed.

An In-Vehicle Warning Information Provision Strategy for V2V-Based Proactive Traffic Safety Management

The availability of vehicle interaction data, which is obtained by an in-vehicle forward collision warning system, including spacing between the leading and the following vehicle and time-to-collision, provides a valuable opportunity to predict crash risks in real time. When this opportunity is combined with connected vehicle technologies including vehicle-to-vehicle wireless communications, it is expected that more effective crash prevention would be achievable by providing predictive warning information as a part of proactive traffic safety management (PTSM). The purpose of this study is to develop a more reliable in-vehicle warning information provision strategy based on the prediction of crash risks using vehicle interaction data. A crash risk prediction model based on a long short-term memory was able to predict the crash risk after 3 seconds with a mean absolute percentage error of 8% using the data for the past 5 seconds. The predicted crash risk data were applied to derive the optimal threshold for triggering in-vehicle warning information, which is the essence of the proposed warning provision strategy. This study defined three indicators to evaluate the reliability of warning information: correct detection rate (CDR), detection failure rate (DFR), and information provision rate (IPR). An exemplar analysis result showed that the optimal threshold to minimize IPR in a situation where CDR and DFR are 100% and 0%, respectively, was identified as 0.69. The proposed methodology that predicts crash risks in real time and provides V2V-based warning information in a more proactive manner is expected to mitigate the crash risk significantly.

Modeling Urban Freeway Rear-End Collision Risk Using Machine Learning Algorithms

A large amount of traffic crash investigations have shown that rear-end collisions are the main type collisions on the freeway. The purpose of this study is to investigate the rear-end collision risk on the freeway. Firstly, a new framework was proposed to develop the rear-end collision probability (RCP) model between two vehicles based on Generalized Pareto Distribution (GPD). Secondly, the freeway rear-end collision risk (F-RCR) was defined as the sum of the rear-end collision probability of each vehicle and divided into three levels which was high, median, and low rear-end collision risk. Then, different machine learning algorithms were used to model F-RCR under the condition of an unbalanced dataset. The result of the RCP model showed continuous change and can identify the dangerous condition quickly compared to the traditional models even when the speed of the leading vehicle is faster than the following vehicle. When the vehicle distribution was unbalanced on road and the speed difference between adjacent lanes and the traffic volume was large, F-RCR will increase. Multi-Layer Perceptron (MLP) was found to be more suitable for modeling F-RCR. The framework provided in this research was transferrable and can be used in the freeway proactive traffic safety management system.