Understanding crash mechanism on urban expressways using high-resolution traffic data

Crash risk analysis for Shanghai urban expressways: A Bayesian semi-parametric modeling approach

With the rapid urbanization during the past three decades in China, the urban expressway network has arisen with large numbers in many cities. However, the traffic conditions are becoming worse and worse. One of the main reasons is that the connected sections of main road and expressway do not have a coordination by the intersection signal and variable speed limits (VSL) on expressway. In this paper, a new cooperation optimization model of VSL on urban expressway and the intersection signal timing are proposed based on the mesoscopic traffic flow model. Its objective is to minimize the total delay and the number of parking. Its essence is adjusting traffic flow on the network in space-time so that the whole network can be in a state of equilibrium. It is effective to the control of local traffic congestion and can avoid the spread of congestion, thus improving traffic safety. Additionally, the optimal VSL strategy is proposed. And then a solution algorithm based on the simultaneous perturbation stochastic approximation (SPSA) is designed for the proposed model. In the end, a practical example of Yanshan overpass in Ji’nan elevated expressway is used to validate the effectiveness of the proposed method. The simulation results are satisfactory. The conclusions are very useful for traffic management department to implement reasonable traffic control strategy.

A real-time crash prediction model for the ramp vicinities of urban expressways

Proper evaluation of ski helmet designs and safety standards should rely on head impact conditions involved in skiing and snowboarding head injuries. To study these impacts, main crash scenarios involving head injuries are numerically replicated. Multibody models of skiers and snowboarders were developed to investigate five common crash scenarios involved in traumatic brain injury: forward and sideways skiing falls, snowboarding backward falls, collisions between users and collisions with obstacles. For each scenario, the influence of crash conditions on head impact (location, speed, linear and rotational accelerations) and risk of injury are evaluated. Crash conditions were initial velocity, user height, position and approach angle, slope steepness, obstacles, and snow stiffness. One thousand one hundred forty-nine crashes were simulated and three significant levels of impact conditions were discriminated over the investigated crash scenarios: 1) the smallest normal-to-slope impact velocities (6 km·h; 22 km·h) and peak linear accelerations (42g; 75g) were obtained during forward and sideways skiing falls; 2) snowboarding backward falls and collisions between users were associated with high normal-to-surface impact velocities (26 km·h; 32 km·h) and head accelerations (80g; 149g) above one published threshold for mild traumatic brain injury but below the pass/fail criteria of helmet standard tests; 3) collisions with obstacles were associated with high normal-to-surface impact velocities (19 km·h; 35 km·h) and the highest head accelerations (626g; 1885g). Current impact conditions of helmet standard evaluations consistently replicate collisions with obstacles, but need to be revised to better reflect other significant crash scenarios leading to traumatic brain injury.

A coordinated traffic control on urban expressways with modified particle swarm optimization

Exploring the impact of a coordinated variable speed limit control on congestion distribution in freeway

Ahstract- Different types of variable speed limit (VSL) controllers have been proposed to improve the traffic mobility at highway bottlenecks. However, the benefit of VSL is sometimes deteriorated by the capacity drop phenomenon introduced by the disordered lane changes at the vicinity of the bottleneck. Our recent results show that by combining the VSL controller with lane change control, the capacity drop can be significantly relieved or avoided at the bottleneck. In this paper, we propose feedback linearization and model predictive control schemes for VSL-actuated highway traffic, where a lane change controller is assumed to be active upstream of the bottleneck. Via simulation studies, we compare the performance and robustness of the proposed controllers with respect to perturbations on traffic demand, model parameters and measurement noise. Results show that both controllers are able to improve the total time spent under different levels of perturbation and noise. Furthermore, feedback linearization VSL can provide better performance than model predictive VSL with negligible computational effort and similar robustness.

Urban expressways connect with local surface/arterial roads tightly through off-ramps, then lead to the increased traffic volume during peak hours, traffic flows near off-ramps usually become chaotic. Therefore traffic flow behavior at off-ramps is critical to improve traffic management of urban expressways with better analysis. Drivers' queue-jump behavior near the off-ramp occurs frequently in the city of developing country, which contributes to the traffic complexity. However, the existing microscopic traffic simulation usually neglected drivers' queuejump behaviors. So, multiple cellular automata based simulation model was built, which specifically targets on queue jump behaviors and the traffic spillback near expressway off-ramp, which the car is contained multiple cellular. Queue-jump rules are incorporated into the car-following and lane changing logics in the model. Using the model, different configurations around off-ramp in combination was simulated. The results show that, easy to recognizable pavement markers is designed to reduce the number of discretionary lane-changes, and multiple cellular cars in simulation reflects the traffic state better.

Variable Speed Limit (VSL) control systems are widely studied as solutions for improving safety and throughput on urban motorways. Machine learning techniques, specifically Reinforcement Learning (RL) methods, are a promising alternative for setting up VSL since they can learn and react to different traffic situations without knowing the explicit model of the motorway dynamics. However, the efficiency of combined RL-VSL is highly related to the class of the used RL algorithm, and description of the managed motorway section in which the RL-VSL agent sets the appropriate speed limits. Currently, there is no existing overview of RL algorithm applications in the domain of VSL. Therefore, a comprehensive survey on the state of the art of RL-VSL is presented. Best practices are summarized, and new viewpoints and future research directions, including an overview of current open research questions are presented.

Variable speed limits (VSLs) and ramp metering (RM) have been proved to be effective alternatives to alleviate the traffic congestions occurred on freeways in many studies. However, no research has so far evaluated the efficiency of the above strategies on urban expressways in Chinese cities, where the early onset breakdown often occurs at the bottlenecks because of different driver behaviour from those reported in overseas. Hence, this study was intended to propose an integrated control strategy of VSL and RM to dynamically postpone the early onset breakdown of on‐ramp bottlenecks based on its occurring probability. The proposed strategy is capable of simultaneously optimising mainline speed and RM rate based on the modified METANET model and the ALINEA model. Meanwhile, an online simulation platform that integrates VISSIM, Visual Basic and MATLAB programs was developed for the purpose of a close‐to‐reality evaluation. Empirical results on urban expressways in Shanghai show that the average travel time and the speed difference were significantly reduced by 8.6 and 41.4%, respectively; the pre‐queue flow at the bottleneck was increased by 14.7%, which indicates that the early onset breakdown could be efficiently postponed or prevented by the use of the proposed control strategy.

A deep reinforcement learning-based intelligent intervention framework for real-time proactive road safety management

Aortic injuries were traditionally thought to be the result of severe frontal crashes. Newer data has suggested other crash types such as nearside crashes may also be important in aortic injury. We hypothesized the implementation of recent safety measures would decrease the incidence of aortic injury associated with fatal motor vehicle crashes. The autopsy reports of all traffic fatalities for motor vehicle occupants in a large urban county for the years 1993 to 2004 were examined. The demographics, impact types, safety measures used, and the presence of any aortic injury were recorded. Trends were evaluated for significance by weighted linear regression. The incidence of aortic injury associated with fatal motor vehicle crashes has remained unchanged during the past 12 years (r = 0.057, p = 0.45). There is a trend toward decreased aortic injuries associated with frontal crashes (r = 0.26, p = 0.089) but no change in aortic injuries associated with nearside or farside crashes (r = 0.053, p = 0.47), when the crash resulted in a fatality. This is despite an increase in seat belt use and increased presence of airbags during the same time period. Despite improved safety measures designed to minimize the occurrence of aortic injuries, the incidence of blunt aortic injury in fatal motor vehicle crashes has not decreased during the past decade. Although not statistically significant, there is a trend toward decreased frontal impacts in fatal motor vehicle crashes associated with aortic injuries. The nearside crash mechanism continues to play a prominent role, and efforts at improving vehicle safety should be focused on crash mechanisms as they relate to aortic injury.

The Highway Capacity Manual prescribes level-of-service (LOS) criteria as a function of traffic density to categorize the operational conditions of both rural and urban freeway sections. This density-based level of service is ideally suited to the assessment of urban freeways when the performance must be optimized to meet high traffic demand. There is, however, some question as to whether density is the appropriate indicator of the quality of service on rural freeways, since drivers may think more in terms of psychological or emotional comfort for freeways, which generally serve long, high-speed trips and rarely experience more than moderate congestion levels. Three specific measures are examined that have at least an intuitive relationship to the concept of driver comfort: ( a) acceleration noise, which is a measure of the physical turbulence in the traffic stream; ( b) number and duration of cruise control applications, which could serve as a general indication of driver convenience or inconvenience; and ( c) percent time spent following, which is already accepted as the basis for determining the level of service on rural two-lane highways. The three candidate measures were estimated for a hypothetical section of rural freeway by simulation, using the CORSIM model. The kinematic relationships of individual vehicles within the traffic stream were estimated by postprocessing the simulation data for each second of operation. All of the measures considered in this study have conceptual appeal. All have produced interesting and potentially usable results with respect to their relationships to traffic volume. Although the simulation results are interesting, further studies focusing on driver opinions, behavior, or field measurements, or all three, would likely be necessary to support the development and recommendation of a specific set of LOS criteria that recognizes the differences between the urban and rural freeway driving environments.

Effective incident detection is essential for emergency response and transportation management. Traditional methods relying on stationary technologies are often costly and provide limited coverage, prompting the exploration of crowdsourced data such as Waze. While Waze offers extensive coverage, its data can be unverified and unreliable. This study aims to identify factors affecting the reliability of Waze alerts and develop a predictive model to distinguish true incidents from false alerts using real-time Waze data, thereby improving emergency response times. Real crash data from the New Jersey Department of Transportation (NJDOT) and crowdsourced data from Waze were matched using the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm to differentiate true and false alerts. A binary logit model was constructed to reveal significant predictors such as time categories around peak hours, road type, report ratings, and crash type. Findings indicate that the likelihood of accurate Waze alerts increases during peak hours, on streets, and with higher report ratings and major crashes. Additionally, multiple machine learning-based predictive models were developed and evaluated to forecast in real time whether Waze alerts correspond to actual incidents. Among those models, the Random Forest model achieved the highest overall accuracy (82.5%) and F1-score (82.8%), and an Area Under the Receiver Operating Characteristic Curve (AUC-ROC) of 0.90, demonstrating its robustness and reliability for real-time incident detection. Gradient Boosting, with an AUC-ROC of 0.90 and Area Under the Precision–Recall Curve (AUC-PR) of 0.90, also performed strongly, particularly excelling at predicting true alerts. The analysis further emphasized the importance of key predictors such as time of day, report ratings, and road type. These findings provide actionable insights for enhancing the accuracy of incident detection and improving the reliability of crowdsourced traffic alerts, supporting more effective traffic management and emergency response systems.

Dynamic pricing has been used to control congestion in mobile wireless networks (MWNs). In dynamic pricing, users are incentivized to time-shift their non-critical usage from the peak hours when the price is high, to off-peak hours when the price is low. The current dynamic pricing (SDP) schemes only consider the congestion level in the call-originating cell and neglect the call-destination cell in the dynamic price computation. The main problem with these schemes is that, when majority of the users in a congested cell are callees, dynamic pricing is ineffective because callers and not callees pay for network services, and resources used by callers and callees are the same for symmetric services. For example, application of dynamic pricing does not deter a callee located in a congested cell from receiving a call, which originates from a caller located in an uncongested cell. To address this problem, we propose a source-destination based dynamic pricing (SDBDP) scheme, which considers congestion levels in both the call-originating and call-destination cells to compute the dynamic price paid by a caller. An analytical model based on Markov decision chain has been developed in this work. Numerical simulations have been carried out in MATLAB for arbitrary caller-callee distributions in a typical day. Call blocking and call dropping probabilities have been used as the performance metrics to compare the current SDP schemes to the proposed SDBDP scheme. Results obtained show that the proposed SDBDP scheme achieves better congestion control than the existing SDP schemes, under arbitrary caller-callee distribution.