2020

Festo Mjogolo, University of North Florida
Thobias Sando, University of North Florida

Various Connected Vehicle (CV) technology applications reported in current literature have the potential to solve challenges that face the transportation sector. Over the last decade, extensive research efforts have been placed on performance evaluation and the benefits of innovative CV applications, and findings indicate that CV technology can effectively mitigate the safety, mobility, and environmental challenges experienced on today’s transportation networks. The majority of previous research has evaluated CV technology through simulation studies.  However, a field study provides a more ideal method of assessing CV technology effectiveness. Field observational studies to validate previous research findings have not been conducted. Therefore, a field study to obtain the actual effectiveness of CV technology was warranted, not only to validate previous findings, but also to add to the body of knowledge surrounding this topic. This study presents a field study evaluation of the effectiveness of CV smartphone technology on a 1.1 mile segment of State Road 121, containing five intersections, in Gainesville, Florida. Field observations were conducted using a CV application, developed by Connected Signals, Inc., that uses a smartphone application, called EnLighten, to communicate intersection information to driver’s smartphone, which serves as a vehicle on-board unit. Traffic operation performance was evaluated using start-up lost time and discharge distribution model. Findings show that the CV smartphone technology improved intersection performance with a reduction in start-up lost time of approximately 86%. There is no impact of CV Technology on the distribution model of position-dependent discharge headways.

Jiali Fu, Statens Vag och Transportforkningsinstitut
Jan Åslund, Linkopings universitet
Markus Rombach, AB Volvo
Ted Samuelsson, AB Volvo
Erik Uhlin, AB Volvo

Autonomous driving offers the construction industry the great opportunity to do the traditional construction process faster, smarter and safer. This paper proposes a methodology for automatically detecting conflict zones in construction operations using autonomous construction vehicles. The methodology employs the vehicles’ dimension and the road geometry to detect the areas with collision risk. Futher, gates are placed on the incoming/outgoing paths of each collision area so that a vehicle can navigate freely in a conflict zone. The obtained information is aimed as input for the fleet control system to regulate the logistics of the vehicles at the conflict zones. Four common types of conflict zones are given in the paper to show that the proposed method is capable to detect the conflict zones in construction environment. Construction sites expand in size and the geometry changes over time. Thus, the proposed method provides an efficient tool to detect the conflict zones in the operating environment.

Fei Ye, University of California, Riverside
Peng Hao, University of California, Riverside
Guoyuan Wu, University of California, Riverside
Kanok Boriboonsomsin, University of California, Riverside
Zhiming Gao, Oak Ridge National Laboratory
Tim LaClair, Oak Ridge National Laboratory
Matthew Barth, University of California, Riverside

Queue length is one of the most essential metrics for indicating traffic conditions at signalized intersections. Most existing studies mainly focus on queue length estimation in a retrieving way which cannot effectively predict the future state of vehicle in queue. This paper proposes an innovative dynamic queue prediction model in a vehicle-infrastructure cooperative way. The proposed dynamic queue predictor can customize the queue prediction using the trajectories of preceding vehicles from radar detection with varying information from loop detectors and connected vehicles (CVs). The dynamic queue prediction accuracy can be further improved with at least one present CV in the same cycle. With the accurate dynamic queue prediction, we present a case study of Queue-Aware Eco-Approach and Departure (QEAD) to provide an online optimal vehicle trajectory considering both the signal status and preceding vehicle's future status in queue. The proposed framework and algorithms are of pragmatic significance and one of its advanced features is the ability to work under low penetration rate of CVs. The test using Next Generation SIMulation (NGSIM) dataset and a case study in VISSIM microscopic simulation show promising results.

Di Kang, University of Minnesota, Twin Cities
Zhexian Li, University of Minnesota, Twin Cities
Michael Levin, University of Minnesota

Autonomous intersection management (AIM) has been widely researched, but previous studies assume that vehicles will follow assigned trajectories precisely. The purpose of this paper is to investigate the safety buffers needed between intersecting vehicles to avoid a collision if a vehicle malfunctions. We optimize vehicle trajectories by deciding the arrival times at each conflict point (point of possible intersection with other vehicles) along each vehicle's trajectory. Because intersecting vehicles rely on the intersection manager (IM) to detect and communicate malfunctions, the reaction time from the IM determines the minimum safety buffer needed. Although a smaller reaction time reduces the safety buffer, it increases the probability that the IM falsely detects a malfunction, instructing vehicles to stop and creating unnecessary delays.This paper develops a mathematical safety buffer for intersecting vehicles, linearizes this time separation, and constructs a combined mixed-integer linear program. A complete protocol is presented and simulated for normal circumstances, emergency circumstances, and recovery circumstances. Sensitivity analyses on various reaction times shows the tradeoff between low reaction times (more false positives) and high reaction times (greater safety buffer). The results indicate that the delay of high-demand scenario is more sensitive than it of low-demand one.

Mehrdad Tajalli, North Carolina State University
Ali Hajbabaie, North Carolina State University

This study presents a mixed-integer nonlinear program for cooperative signal timing and trajectory optimization at signalized intersections with mixed traffic of connected and automated vehicles (CAVs) and connected human-driven vehicles (CHVs). The proposed program optimizes the signal timing parameters and CAV trajectories based on data from CAVs and CHVs. The trajectories of CHVs are estimated based on Helly’s car following concept. We have integrated traffic signal indications in the car following formulations as well. The resulting optimization problem is a mixed-integer nonlinear program, is intractable, and cannot be solved to the optimal level for complex intersections. We linearize the nonlinear constraints and reformulated the problem with a tight convex hull of the mixed-integer solutions. Then, we develop a Lagrangina relaxation technique to decomposes the optimization problem into several lane-level optimization sub-problems. Hence, a unique controller optimizes trajectories of all vehicle on a lane and the signal timing parameters associated with the same lane. This setting will allow finding near-optimal solutions with small duality gap for complex intersections with different movements, lane groups, and high demand levels. The complexity of the problem is further reduced by embedding the proposed approach in a receding horizon control. Case study results show that the proposed methodology finds near-optimal signal timing plans and trajectory of CAVs efficiently with at most 0.1% Lagrangian duality gap. The results also show that increasing the penetration rate of CAVs reduces the average travel time at signalized intersections by 19%-61% in comparison with optimized fully-actuated signal timing plans.

Xishun Liao, University of California, Riverside
David Oswald, University of California, Riverside
Ziran Wang, Toyota Motor North America
Guoyuan Wu, University of California, Riverside
Kanok Boriboonsomsin, University of California, Riverside
Matthew Barth, University of California, Riverside
Kyungtae Han, Toyota Motor North America
BaekGyu Kim, Toyota Motor North America
Prashant Tiwari, Toyota Motor North America

ABSTRACT Ramp merging has been a popular research topic in recent decades, since they are often associated with dangerous human behaviors and inefficient traffic movements. With emerging connected and automated vehicle (CAV) technology, a cooperative ramp merging system has been developed with vehicle-to-cloud (V2C) communications, enabling safer and more efficient merging behaviors. In this paper, intelligent speed assistance is utilized to show the suggested speed to the driver of a CAV, allowing him/her to perform a cooperative merge with another CAV from the other lane using V2C communications. Field experiments with different settings have been conducted in this study to evaluate the performance of our previously proposed cooperative ramp merging system and human tracking error prediction model. The first “virtual” cooperative ramp merging field experiment shows the error prediction model is capable of compensating the speed tracking errors generated by a human driver, with a reduction of 20% on the speed error variance compared to the baseline. The second two-vehicle cooperative ramp merging field experiment shows the cooperative ramp merging algorithm can smooth the speed trajectories of the ramp merging vehicle and also validates the concept under the V2C communications framework.

Meng Li, Southeast University
Zhibin Li, Southeast University
Ling Zhao, Southeast University
Yang Zhou, University of Wisconsin, Madison

This paper proposes a cooperative jam-absorption driving (CJAD) strategy on multi-lane highways to mitigate traffic oscillations. The strategy is based on a prediction of the formation and propagation of oscillations and is developed for low penetration rates of connected and autonomous vehicles (CAVs). It considers lane changes and interactions between different lanes and aim at eliminating jams simultaneously on all lanes. When an oscillation is detected, the strategy firstly controls a small number of CAVs in different lanes to arrive a same longitudinal position to form a barrier that prevents overtaking. Then the strategy slows down these CAVs simultaneously to a designed speed (i.e. absorption speed) to keep a big gap with downstream oscillation until it disappears. The speed is designed obeying a rule that the CAVs catch up with the preceding vehicles upon the oscillation dissipates. A parsimonious oscillation prediction model was adopted to estimate the time-space ending point of oscillation and decide the CAVs’ absorption speed. Two scenarios with one oscillation and a series of oscillations were constructed to test the performance of proposed strategy based on the Simulation of Urban Mobility (SUMO). The results show that the proposed multi-lane CJAD strategy is able to absorb traffic oscillations, improve traffic safety, as well as reduce traffic disturbance.

Hiba Nassereddine, University of Wisconsin, Madison
Kelvin Santiago-Chaparro, University of Wisconsin, Madison
Jonathan Riehl, University of Wisconsin, Madison
David Noyce, University of Wisconsin, Madison

Research was conducted to evaluate the effectiveness of a warning system that communicates the presence of a potential non-connected red-light-running vehicle to the driver of a connected vehicle. The warning system was built into a full-scale driving simulator and included a combination of an auditory cue accompanied by a visual message on the windshield as head-up-display. Twenty participants were recruited for the experiment and were exposed to an imminent collision scenario with a non-connected red-light-running vehicle. Participants were randomly placed in a control group and three treatment groups and their response to the scenario was studied. For the control group, the warning system was activated at the intersection stop bar pavement marking, whereas for the treatment groups, the warning system was activated at 50 ft, 100 ft, and 150 ft from the stop bar. The reaction time to the warning system was collected and Kruskal-Wallis test was used to investigate the significance in the differences of means. The results showed a statistically significant difference in the mean of reaction time between the different groups. Mann-Whitney Wilcoxon test was used to compare between two groups. Drivers reduced their speeds for an average of 2.2, 2.2, 2.6, and 3.2 sec when the warning system was activated at the stop bar and 50, 100, and 150 ft before the stop bar, respectively. Participants came to a complete stop for 29.7%, 17.5%, 29.3%, and 47.7% of the events at the stop bar and at 50, 100, and 150 ft before the stop bar, respectively.

Xing Fu, University of Alabama
Qifan Nie, University of Alabama
Jun Liu, University of Alabama
Asad J. Khattak, University of Tennessee
Alexander Hainen, University of Alabama
Shashi Nambisan, University of Nevada, Las Vegas

Connected and automated vehicles (CAVs) are expected to change the way we travel. Before both the vehicles and infrastructures are fully automated, users of CAVs are required to respond appropriately to any adverse on-road conditions or malfunction that may prevent the autonomous driving system from reliably sustaining the dynamic driving task performance. The objective of this study is to construct spatiotemporal driving volatility profiles to help CAVs or drivers identify the potential hazards in the existing transportation network and make proactive driving decisions.  The volatility profiles are constructed based on the historical traffic dynamics, varying spatially and temporally in the network. For demonstration, this study exploited the Basic Safety Messages datasets from Safety Pilot Model Development program in Ann Arbor, Michigan. The driving volatility is a measure to reflect the variability of driving performance, which is often used to show a vehicle or driver’s performance on road. This study extends the concept to capture the driving dynamics as a performance of the transportation network. This study also matched the driving volatility to the spatial and temporal occurrence of historical traffic crashes. Modeling results showed the volatility is significantly related to safety outcomes; therefore, the driving volatility profiles can be useful in terms of informing CAVs and drivers of potential on-road hazards and assisting in making proactive driving decisions. Further, the results offer implications for potential upgrades of the transportation infrastructure for full automation in the future.

Burak Sen, Connected Wise LLC
Enes Karaaslan, Connected Wise LLC
Haluk Laman, University of Central Florida
Tolga Ercan, Connected Wise LLC
James Pol, Federal Highway Administration (FHWA)

Connected automated vehicles (CAV) are expected to revolutionize the way goods and people move, further increasing the overall efficiency of transportation systems management and operations. However, proper intelligent transportation system infrastructure, compatible with wireless communication technologies, is crucial for benefiting from the CAV technology. However, providing and maintaining such an infrastructure is a challenging task, particularly when it comes to rural areas. This study proposes and presents the field test of the implementation of machine-readable, smart infrastructure-to-vehicle (I2V) signs that are used to relay the same I2V application messages that are sent by roadside units. The objective of the study is to demonstrate the efficiency of a low-cost solution to the challenges encountered in rural areas in terms of the implementation of the CAV technology. For this purpose, three I2V communication scenarios, namely MapData message, Traveler Information Message, and Red-Light Violation Warning message, are presented. Other I2V messages that can be relayed in accordance with the relevant standards using the proposed approach are also shown. The field tests show promising results for faster deployment of the CAV technology in rural areas as well as urban areas. The proposed approach is found to be robust enough for reliable deployment. One of the main advantages of the proposed approach is that it is not vulnerable to electromagnetic interference or cyber-security attacks, and is capable of robustly functioning regardless of wireless communication technology (e.g. dedicated short-range communication and cellular communication such as 5G-LTE).

Rasool Mohebifard, North Carolina State University
Ali Hajbabaie, North Carolina State University

This paper presents a formulation for trajectory optimization of Connected and Automated Vehicles (CAVs) in roundabouts as a non-linear and non-convex optimization program and presents a customized solution technique to solve it efficiently. The program minimizes the total travel time of CAVs at a roundabout and includes vehicle dynamics and collision-avoidance constraints with explicit representation of vehicles paths. Because the program is non-convex and non-linear, it cannot be solved with conventional optimization techniques. Thus, we developed a customized solution technique that convexifies the collision-avoidance constraint and employs the Alternating Direction Method of Multipliers to decompose the convexified problem into two sub-problems, i.e., Sub-problems 1 and 2. We show that Sub-problem 1 only includes vehicle dynamics constraints, and Sub-problem 2 projects the solutions of Sub-problem 1 onto a collision-free region. Sub-problem 1 is further transformed into a quadratic program by redefining its decision variables along vehicles paths. This transformation allows distributing Sub-problem 1 among vehicles such that each vehicle solves an optimization program including its own objective function and constraints. Moreover, we show that the iterations between Sub-problems 1 and 2 converge to the optimal solutions of the convexified problem. The solution technique is applied to a case study roundabout with different demand profiles. The results showed that the trajectory optimization reduced the total travel time and average delay of vehicles respectively by 15.43% to 50.99% and 85.93% to 95.46% compared to a scenario where a conventional car-following model determined the trajectories of vehicles.

Eric Adomah, University of Wyoming
Guangchuan Yang, North Carolina State University
Mohamed Ahmed, University of Wyoming

Reducing crash risk in work zones is a paramount goal for the Federal Highway Administration. With recent strides in improving work zone safety using countermeasures including several Intelligent Transportation Systems, the number and severity of crashes at work zones are still notable. Similarly, an exogenous factor, such as fog can increase crash risk. The advent of CV technology is poised to influence driver behavior and hence reduce crash risk positively. The study which is premised on the Wyoming DOT Connected Vehicle Pilot Program sort to provide preliminary insight into how the behavioral adjustments of connected trucks to CV warnings in a rural work zone during low visibility presented by fog conditions will affect work zone safety using microsimulation modeling. Findings from the study indicated that there was more homogeneity in speeds of vehicles in all segments of the work zone at 30% Market Penetration Rate. Most of the work zone conflicts during foggy conditions occurred at the advance warning area of the work zone. In terms of surrogate safety measures, the advance area of the work zone consisted of more than 50% of the total conflicts in all scenarios. Also, at 30% MPR, there was a significant reduction in conflicts by 21% from the base scenario. The results suggested that the minimum Market Penetration Rate at which we gain significant safety benefits is when all trucks in the rural work zone are connected (i.e.,30% MPR).

Asad J. Khattak, University of Tennessee
Numan Ahmad, University of Tennessee
Behram Wali, Urban Design 4 Health, Inc.

The transportation system of the future is anticipated to integrate automation, connectivity, electrification, and sharing of rides and vehicles—a concept known as ACES. Driven by the growing computational power, ubiquity of sensors, big data, and Artificial Intelligence, the public and private sectors have new opportunities to improve the quality of lives through greater mobility, safety, energy efficiency, and cleaner environments. For example, automation can provide safer mobility; electrification (where appropriate) can reduce dependence on fossil fuels, and sharing of vehicles can improve access. The value embedded in ACES can be harnessed better by understanding how they may be adopted by consumers. This paper seeks to understand the behavioral impacts of ACES by identifying the characteristics of early adopters and their willingness to purchase/use services offered by ACES. Consumers’ affinity toward automation including self-driving vehicles, is explored in this study. Specifically, using a survey of California consumers (N=3,429), the willingness to purchase automated vehicles, the ownership of electric vehicles and households’ participation in car-sharing programs is analyzed together using a path analytic framework. Households that currently own electric vehicles, and participate in car-share and ride-share programs are also more willing to purchase automated vehicles. Within the methodological limitations of survey research for consumer preferences, the study highlights the consumers’ concerns inherent in transitioning to ACES. Overall, the wider adoption of automation, electrification, and car-sharing can be enhanced if consumers’ preferences and concerns are appropriately addressed in the future.

Jackeline Rios-Torres, Oak Ridge National Laboratory
Jihun Han, Argonne National Laboratory
Ramin Arvin, University of Tennessee, Knoxville
Asad J. Khattak, University of Tennessee

Coordination control systems for vehicles merging at on-ramps can potentially improve safety, mobility and energy consumption. We present and evaluate two optimal coordination systems (OCSs), one minimizing acceleration-based cost and other minimizing fuel consumption. They allow connected and automated vehicles (CAVs) to merge onto a highway safely and smoothly. In the proposed OCSs, a high-level controller coordinate CAVs to reach the merge area based on a first-in-first-out queue, while a low-level controller computes the optimal reference speed to be followed. The impact of the two OCSs on safety and fuel consumption is assessed through microscopic traffic simulation, and insigths on potential mobility impacts are also obtained. Simulation results show that the OCSs can reduce fuel consumption and the total time vehicles take to travel across the merging scenario under study. This is when the performance is compared between the coordinated scenarios and a baseline uncoordinated scenario. Notably, fuel consumption savings remain very close for the two OCSs. This is an important finding given the potential for real time implementation of the optimal acceleration solution compared to the optimal fuel consumption solution. Moreover, using the number of conflicts and driving volatility as safety indicators, the coordinated scenarios outperform the baseline scenario, reducing unsafe driving meneuvers.  Overall, this study provides technical understanding of the optimal coordination problem to enhance operation of CAVs. Hence, it provides insigths on potential future controls and methods that promote cooperative, efficient operation to inform public and private sector decision-making in deploying connectivity-based applications for Automated Vehicle technologies.

Baris Cogan, Technische Universitat Munchen
Eftychios Papapanagiotou, Technische Universitat Munchen
Tobias Schendzielorz, Schlothauer & Wauer GmbH
Fritz Busch, Technische Universitat Munchen
Sasan Amini, Technische Universitat Munchen

Current Urban Traffic Control (UTC) systems rely on infrastructure-based detectors. The emerging technologies such as Connected Vehicles (CV) and new available data sources such as aggregated Floating Car Data (FCD) bring new possibilities for improving signal control, while reducing the reliance on loop detectors. New approaches are needed in order to integrate these data sources into existing signal control strategies for the upcoming transition phase with low penetration rate of connectivity. This paper presents a methodology that utilizes CV and aggregated FCD in existing traffic-actuated signal control for investigating their immediate potential in current systems without the need for a new control method. A traffic state estimation module based on the Extended Kalman filter is used in order to fuse CV and aggregated FCD for estimating traffic parameters (arrival rate, departure rate and queue length) necessary for the signal control operation without loop detectors. The results of the estimation show that the fused estimations outperform the sole CV measurements at all penetration rates. An actuation module uses the estimation and the CV presence to replace the loop detectors. The presented methodology is able to replace tailback, time gap and phase call detectors. The replacement of the tailback detector seems to have the most potential based on the accurate estimation of the queue length and can lead to reduced delays for oversaturated signalized approaches. However, the phase call and time gap detectors are naturally less accurate in low penetration rates and low volumes and can lead to increased delays.

Nigel Williams, University of California, Riverside
Alexander Vu, University of California, Riverside
Guoyuan Wu, University of California, Riverside
Matthew Barth, University of California, Riverside
Kun Zhou, University of California, Berkeley

Connected Vehicle (CV) technology has the potential to greatly improve the safety, mobility, and environmental sustainability of traffic. Many CV applications require the vehicle position as input, which is typically provided by a GNSS receiver embedded in an aftermarket Dedicated Short Range Communications (DSRC) unit. Although a large number of those applications (e.g., Intersection Movement Assist) require vehicle positioning to have lane-level accuracy, it has been shown that the type of positioning system typically used by CVs currently cannot provide consistent lane-level accuracy, even under open-sky conditions. In order to address this gap, we have evaluated an enhanced positioning system which adds little, if any, to the cost of the CV. It consists of a single-frequency RTK-capable GNSS receiver onboard the vehicle, which utilizes Radio Technical Commission for Maritime Services (RTCM) differential corrections transmitted over DSRC by the roadside infrastructure. Tests on a moving vehicle show that this system could provide lane-level accuracy over 95% of the time in open-sky conditions. These tests also show DSRC to be an effective means of disseminating RTCM corrections, given the intersection spacings and communication ranges in the test. However, neither RTCM nor the more commonly used Space-Based Augmentation System (SBAS) differential corrections appeared to improve the positioning accuracy of GNSS in urban canyons.

Chengyuan Ma, Tongji University
Chunhui Yu, Tongji University
Hailun Liang, Tongji University
Xiaoguang Yang, Tongji University

Trajectory planning for connected and automated vehicles (CAVs) has the potential to improve operational efficiency and vehicle fuel economy in traffic systems. Despite abundant studies on this research area, most of them only consider trajectory planning in the longitudinal dimension or assume the fully CAV environment. This study proposes an approach to the distributed optimization of CAV trajectories at an isolated signalized intersection under the mixed traffic environment, which consists of connected and human-driven vehicles (CHVs) and CAVs. A bi-level model is formulated to optimize the trajectory of a CAV in both the longitudinal and latitudinal dimensions given signal timings, predicted trajectories of CHVs, and planned trajectories of CAVs. The upper-level model optimizes lane-changing behaviors, and the lower-level model optimizes longitudinal trajectories based on the lane-changing strategies from the upper-level model. Minimization of vehicle delay, fuel consumption, and lane-changing costs are adopted as the objective function. A branch and bound algorithm is applied to solve the bi-level optimization model. An implementation procedure is designed for the application of the proposed model to time-varying traffic condition. The sequence of trajectory optimization for newly arrived CAVs is determined by “first-come-first-service” policy. Numerical studies validate the advantages of the proposed model. For one CAV, fuel economy can be improved by up to 35% and average delay can be reduced by around 2 s. The performance of the proposed model improves remarkably with the increasing CAV penetration rate in terms of vehicle delay, fuel economy, and vehicle throughput.

Zulqarnain Khattak, Oak Ridge National Laboratory
Brian Smith, University of Virginia
Michael Fontaine, Virginia Transportation Research Council
Jiaqi Ma, University of California, Los Angeles
Asad J. Khattak, University of Tennessee

Lane control signals (LCS) mounted on overhead gantries have been used to provide merge advisories to human drivers when lanes are closed due to incidents or work zones. The development of connected and automated vehicles (CAVs) provides opportunities to develop applications for sending advisory messages directly to CAVs, which would substantially reduce infrastructure costs over current LCS approaches. The paper developed an enhanced CAV-enabled LCS application to overcome limitations of infrastructure-based systems and improve operations and safety in a mixed traffic environment consisting of CAVs and human driven vehicles. Impacts of this approach on operations and safety are compared to traditional infrastructure-based LCS using a calibrated baseline simulation model of a site where traditional LCS was present. The simulation results showed that starting at a lower penetration of 20% CAVs, an average throughput improvement of 6.38%, 5.72%, and 5% were observed for 1s, 1.5s, and 2s headways. Conflicts were reduced by an average of 61% and 69% for rear end and lane change conflicts, respectively. Likewise, the CAV-enabled LCS was found to reduce volatility, represented by variation in acceleration and deceleration, by an average of 18.2% and 17.6% under varying penetrations of CAVs. Volatility represents the variation in driving behavior over an instantaneous period of driving, so this reduction in volatility serves as a positive safety surrogate indicator for this new application as well.

Yu Zhang, Tongji University
Yu Bai, Tongji University
Jia Hu, Tongji University
Meng Wang, Technische Universiteit Delft

In the past, most Connected and Automated Vehicles (CAVs) applications are developed separately, hence, is usually functional one at a time. This causes concerns over safety, stability and robustness of cooperative automation, as switching between different applications potentially lead to system failure. In order to integrate different applications into one, this research developed a generic optimal control framework for various cooperative automation applications. A model predictive control based problem formulation is proposed. A solution method based on dynamic programing is designed. The convergence of the proposed solution method is proved and convergence speed is found to be faster than classic Newton method. To showcase the proposed framework, example formulations for four cooperative automation applications are developed. Including adaptive cruise control (ACC), multi-anticipating adaptive cruise control (MACC), cooperative adaptive cruise control (CACC) and cooperative lane change (CLC). The proposed solution method is applied to solve for the four applications and evaluated against PMP method. Results show that computation time of the proposed method is 7.8 - 8.2 milliseconds, which is 84% faster than iPMP method.

Ruochen Hao, Tongji University
Chunhui Yu, Tongji University
Wanjing Ma, Tongji University
Ling Wang, Tongji University
Xiaodong Zhu, China Highway Engineering Consultants Corporation

Existing traffic control systems mainly rely on infrastructure-based detectors (e.g., loop detectors), which have high installation and maintenance costs. The data collected by infrastructure-based detectors only provide the information of traffic flows at an aggregate level (e.g., traffic volumes). With the development of connected vehicles (CVs), vehicle trajectory data become available and cheap. The advantages of trajectory data lie in that they track vehicle states (e.g., positions and velocities) at an individual level in both spatial and temporal dimensions. Vehicle trajectory data have been used in traffic control and management as well as traffic estimation and prediction. However, the challenge is the low penetration of CVs. Reconstruction of the trajectories of unconnected vehicles/regular vehicles (RVs) provides a potential solution. This study proposes an approach to combine loop detector data and CV trajectory data to reconstruct RV trajectories under the partially CV environment. A linear car-following model is selected to capture the driving behaviors of CVs and RVs, which is calibrated using the Next Generation Simulation (NGSIM) data. The differences between the driving behaviors of different vehicles are differentiated. Maximum likelihood estimation (MLE) is applied to reconstruct the trajectories of RVs between consecutive CVs in a lane based on loop detector data and CV trajectory data. The formulated optimization model proves to be convex and can be easily solved by existing algorithms. The numerical study validates the advantages of the proposed model over three benchmark models in terms of estimation accuracy.

Xiangguo Liu, Northwestern University
Guangchen Zhao, University of Maryland, College Park
Neda Masoud, University of Michigan
Qi Zhu, Northwestern University

Autonomy and connectivity are considered to be among the most promising technologies to improve safety, mobility, and fuel consumption in transportation systems. Some of the fuel efficiency benefits of connected and automated vehicles (CAVs) can be realized through platooning. A platoon is a virtual train of CAVs who travel together following the platoon head, with small gaps between them. In this paper, we devise an optimal control-based trajectory model that can provide fuel-efficient trajectories for the subject vehicle and can incorporate platooning. We embed this trajectory planning model in a simulation framework to quantify its fuel efficiency benefits in a dynamic traffic stream. Furthermore, we perform extensive numerical experiments to investigate whether the vehicles upstream of the subject vehicle may also experience second-hand fuel efficiency benefits.

Moahd Alghuson, Southern Methodist University
Khaled Abdelghany, Southern Methodist University
Ahmed Hassan, Cairo University

This paper conceptualizes a score-based traffic law enforcement and network management system (SLEM) which leverages the connected vehicle (CV) technology. The system assigns a real-time score for each driver to reflect her/his driving performance. Different from current systems that issue tickets to violating drivers, the proposed system adopts a personalized route guidance strategy that tends to favor high-scoring drivers by guiding them to less congested routes on the expense of low-scoring drivers who are directed to alternative slower routes. The strategy is designed such that it shifts the traffic distribution pattern in the network from the undesirable user equilibrium (UE) pattern to the system optimal (SO) pattern. A mathematical formulation in the form of bi-level formulation and an efficient solution algorithm are developed to solve the problem. A set of experiments are conducted to evaluate the performance of SLEM considering different operation scenarios.

Md Sharikur Rahman, HDR
Mohamed Abdel-Aty, University of Central Florida

Connected vehicles’ (CV) technologies are expected to have a great potential on traffic safety and operation in the transportation road network. However, the most common problem impeding the CVs popularization is market penetration rate (MPR). Therefore, the study of market penetration rates is worthwhile in the CV transition period. This paper aims to investigate traffic safety benefits of CV platoons by comparing the implementation of dedicated lane CV platoons and all lanes CV platoons (with same MPR) over non-CV scenario using microsimulation (VISSIM). This study applied the concept of CV platooning on 22 miles section of a congested expressway in Florida to improve traffic safety exploring the appropriate utilization of MPRs. To regulate the driving behavior of CV platoons, the Intelligent Driver Model (IDM) along with platooning algorithms were implemented. Three joining strategies with high-level control algorithm were developed such as rear join, front join, and cut-in joint to implement the CV platooning for both dedicated lane CV platoons and all lanes CV platoons. The crash risks of the two CV platoon’s scenarios were analyzed based on five surrogate measures of safety: standard deviation of speed, time exposed time-to-collision (TET), time integrated time-to-collision (TIT), time exposed rear-end crash risk index (TERCRI), and sideswipe crash risk (SSCR). The results suggest that both dedicated lane CV platoons and all lanes CV platoons reduce crash risk significantly in terms of the five surrogate measures of safety. The dedicated lane CV platoons significantly outperformed all lanes CV platoons with the same MPR.

Jian Gong, Southeast University
Jianhua Guo, Southeast University
Jinde Cao, Southeast University
Yun Wei, Beijing Urban Construction Design & Development Group Co., Limited
Wei Huang, Southeast University

This paper presents a distributed coordination scheme for connected automated vehicles (CAVs) in a non-signalized intersection. A novel framework of cooperative intersection control is proposed to divide an intersection area into three sections, including the free zone, the platoon zone and the control zone. To avoid collision of each pair of the conflict platoons approaching from different directions, a platoon-based coordination strategy is designed by scheduling the arrival time of each leading vehicle in different platoons. In order to enable the following vehicles to track the trajectory of the leading vehicle, as well as guarantee the desired distance between any two adjacent vehicles, distributed platoon controllers are designed for all the following vehicles. On this basis, the vehicular platoon is taken as a whole to be coordinated with the help of intersection coordination unit (ICU). Specially, to avoid a collision at an intersection and improve transportation efficiency and fuel economy, the travel time and fuel consumption are considered as the performance index comprehensively. Thus, the optimal control problem of the leading vehicle is formulated with the variable constraints derived from coordination with other approaching platoons from different directions. The Pontryagin Minimum Principle (PMP) and Phase-plane method are applied to find the solution of the optimal control problem. Numerical simulations show the effectiveness of the proposed scheme.