2021

Ning Wu, Ruhr-Universitat, Bochum
Werner Brilon, Ruhr University, Bochum

For the estimation of capacities and traffic performance at Two-Way-Stop-Controlled (TWSC) intersections, investigations have been conducted by many researchers. The results of these investigations are incorporated in highway capacity manuals like HCM or HBS. Although the underlying methodologies are similar, there are two major differences in the current HBS 2015 and HCM 2016. That is, a) the procedure for the impedance factor for movements of rank 4 and b) the procedure for estimating the capacity of shared short-lanes both for minor and major movements. In the HBS 2015, new developments have been account for and the accuracy of capacity and traffic quality estimations has been significantly improved. In the HCM 2016, the two procedures are not updated up to now. Therefore, the replacement of the two procedures in the current HCM 2016 are recommendable. Both in HCM 2016 and HBS 2015, the procedures for calculating delays at shared lanes or shared short-lanes are inaccurate and they should be updated as well. In most cases, the delays are significantly underestimated. Recently, the authors have developed a new methodology dealing with this problem. This methodology can be easily incorporated into future versions of the HBS and HCM. In this paper, the theoretical backgrounds of the three new methods are presented and major results are summarized. Compared to the HCM 2016, the advantages of the new developments are highlighted. As a recommendation, three corresponding procedures for capacity and delay estimations are given for potential use in a future version of the HCM.

Pinakin Patel, Sardar Vallabhbhai National Institute of Technology
Dr. Ashish Dhamaniya (adhamaniya@gmail.com), Sardar Vallabhbhai National Institute of Technology

This study proposes the time occupancy approach to estimate the passenger car units (PCUs) at urban signalized intersections with different level of interaction between the vehicles under saturated green time condition. The study shows variation in PCUs with varied traffic and geometric conditions. Traffic data have been collected through videography technique at signalized intersections in three metropolitan cities in India. Traffic flow discharge and clearance time of different vehicular categories have been extracted from the video during the saturated green time. The observed range of dynamic PCU values for 2W, 3W, CB, LCV and HV are 0.12-0.32, 0.45-0.80, 1.40-1.80, 1.60-2.20 and 3.50-6.50 respectively. To address the variation in individual vehicle’s PCU with traffic compositions and flow rate, regression-based PCU models are developed for each category of vehicles. The model analysis shows that traffic compositions and flow rate are significantly affecting the PCU values. The PCU is a complex parameter and required number of field attributes and hence to overcome the complexity to estimate PCUs, a concept of Flow Equivalency Factor (FEF) has been proposed based on the estimated PCUs. The FEF can directly be used to convert the mixed motorized vehicular flow into an equivalent standard passenger car flow without actually making use of individual PCU factor of different vehicles. All the developed models are validated for field conditions and results found promising with field data. The developed approach can be used effectively for developing countries with a mixed traffic stream.

Zihang Wei, Texas A&M University, College Station
Yunlong Zhang, Texas A&M University, College Station
Xiaoyu Sky Guo (xiaoyuguo@tamu.edu), Texas A&M University, College Station
Xin Zhang, Texas A&M University, College Station

The through movement capacity is an essential factor used to reflect intersection performance, especially for signalized intersections where a large proportion of vehicle demand is through moving vehicles. For intersections with short left-turn bays, left-turn spillback and blockage can occur frequently. Left-turn spillback is considered to affect through movement capacity and blockage to the entrance of left-turn bay is considered to decrease left-turn capacity. Previous studies mainly focused on estimating through movement capacity under a lagging protected only left-turn (lagging POLT) signal setting because left-turn spillback is more likely to happen under this timing strategy. However, no comprehensive study was conducted to assess the impact on through capacity under all commonly used left-turn signal settings. The existing through movement capacity model proposed by HCM does not consider the existence of spillback and blockage, so it cannot estimate the through movement capacity with short left-turn bay accurately. In this paper, through movement capacity models based on probabilistic modeling of spillback and blockage scenarios are established under four different signal settings. Apart from leading POLT and lagging POLT, protected plus permitted (leading PPLT) and permitted plus protected (lagging PPLT) are also included. Through microscopic simulations, the proposed models are compared with models proposed by other studies and in the HCM. The comparisons results demonstrate that the proposed models have significant advantages over all the other models and have high accuracy in all signal settings. Each proposed model for a given signal setting maintains consistent accuracy across various left-turn bay lengths.

Rui Yue, University of Nevada, Reno
Guangchuan Yang, North Carolina State University
Yichen Zheng, Beijing Nebula Link Technology Co., Ltd.
Zong Tian, University of Nevada, Reno

Two-way stop-controlled (TWSC) intersections have been applied extensively in the U.S. and other parts of the world when traffic signal control is not warranted. However, in reality it was found that when a major road vehicle yields to crossing pedestrians, minor road traffic could utilize this extra gap, which tends to improve the capacity of some minor vehicle movements. The current capacity modeling methods did not take into account the effects of the pedestrian crossing on minor road capacity. This paper proposed an analytical model to quantify the increased capacity of minor street movements contributed by minor street pedestrian crossings and validated the model using both field data collected at a three-leg TWSC intersection and through the stochastic simulation method. A sensitivity analysis was performed to reveal the impacts of various factors on minor road capacity. In general, it was found that minor road left turn capacity at the study intersection was positively correlated to pedestrian crossing volume, yielding rate, and pedestrian crossing time. Nonetheless, the increased capacity tended to be restricted by the initial capacity and the maximum capacity of the minor road, indicating that pedestrians can only bring in conditional capacity increases. Modeling results showed that under a relatively heavier conflict traffic volume condition, the effect of the pedestrian crossing on minor road increased capacity was more significant.

Muhammed Hafiz A.N., Trivandrum
Aneesh Babu, Trivandrum
Anusha P. (anushanair1@yahoo.co.in), Trivandrum

Intersections are the most common places in a road network where vehicles face significant delay while traveling. Traffic in developing nations such as India is highly heterogeneous in nature with poor lane discipline. As a consequence, conventional delay estimation approaches developed for homogeneous counterpart tend to provide erroneous results. This paper compares the delays estimated by 2010 version of the Highway Capacity Manual (HCM), Indian Highway Capacity Manual (Indo-HCM, 2017) and a modified HCM model developed by Saha et.al. (2017) for mixed traffic conditions in addition to the delays estimated from the Radio Frequency Identifier (RFID) sensors. The delays were estimated at a pre-timed signalized intersection under mixed traffic conditions. Based on the results, it was observed that the delay estimated by the Modified HCM model and the delays determined from RFID sensors best replicates the actual delay computed using Simpson’s one-third rule. Mean Absolute Percentage Errors (MAPE) were observed to be of 10.8% and 10.4% respectively indicating a good estimation of delay by these methods. Since the delay estimation from models require numerous traffic variables which are often tedious to obtain from the field under mixed traffic conditions, the delays estimated from sensor data in an automated manner will be helpful in developing Advanced Traffic Management Systems (ATMS) strategies for Intelligent Transportation Systems (ITS) application in real time under mixed traffic conditions.

Ahmed Elawady, University of Sharjah
Muamer Abuzwidah, University of Sharjah
Waleed Zeiada, University of Sharjah

Technology has advanced rapidly in the past few decades, Connected Vehicles (CVs) is one of these technologies. The main goal of using CVs is to improve traffic safety; however they are also expected to reduce traffic delay and delay costs at intersections. This paper aims to examine the impact of using CVs technology on traffic delay at signalized intersections in the emirate of Sharjah, UAE. The study area was chosen to be the intersection between Al-khan Street and Second Industrial Street. Average delay and stop delay were discovered using VISSIM traffic micro simulator. Behaviors of human-driven vehicles and CVs were modeled in the VISSIM’s Wiedemann 99 car-following model. Different scenarios were simulated in VISSIM based on the penetration rate of CVs at the intersection. The penetration rates were at Zero percent as a base case, 25 percent, 50 percent, 75 percent, and 100 percent. The results showed that CVs improved intersection capacity and reduced traffic delay significantly with the increase of penetration rates. The CVs reduced traffic delay by 4 to 10 percent with the penetration rates between 25 and 100 percent at a ninety-five percent confidence level. For the Stop delay, the reductions were between 3 and 8 percent, which are evident with the 100 percent CVs penetration rate at a ninety-five percent confidence level. As for traffic delay cost, CVs showed a reduction of the traffic delay cost by at least 12 million AED a year by each intersection when the intersection is fully functioned with Connected Vehicles.

Fabio Sasahara (fsasahara@ufl.edu), University of Florida
Ageliki Elefteriadou, University of Florida

The Highway Capacity Manual (HCM), currently in the Sixth Edition, has been for many decades the major reference for agencies and practitioners to evaluate capacity and quality of service on roads. However, it does not address cases when a queue spills back into ramp terminals. This paper proposes a new method for retiming signalized intersections affected by on-ramp queue spillback in order to minimize its impacts on the overall intersection performance. This new procedure evaluates whether  spillback occurs, and then estimates the resulting loss of capacity for the intersection’s affected movements. This is used to estimate the wasted effective green time, which can be re-allocated to other movements without affecting the throughput of movements affected by queue spillback. The proposed method connects the HCM methodologies of signalized intersections and freeway facilities since their operations are interdependent and need to be analyzed collectively to determine whether queue spillback is expected. On-ramp arrivals are determined based on the signalized intersection operation and on-ramp departures are obtained using freeway facilities methods. Next, queue profiles at the on-ramp are constructed on a cycle-by-cycle basis to determine the time of spillback into the ramp terminal. Results show that delay in other movements can be reduced, contributing to mitigating the impacts of on-ramp congestion in arterial streets.

Nemanja Dobrota (ned47@pitt.edu), University of Pittsburgh
Aleksandar Stevanovic, University of Pittsburgh
Nikola Mitrovic, Siemens Mobility

Traffic signal performance measurement plays an essential role in all stages of the system development process. In recent years, the concept of Automated Traffic Signal Performance Measures (ATSPM) as a means to improve signal operations was proposed. Vehicle delay is one of the fundamental performance measures that represent an inevitable part of traffic signal optimization procedures. Over time, the process of delay estimation shifted from various analytical models towards the use of real-time methods. Each delay model is specific in terms of its inputs and outputs. Analytical models provide information about the quality of progression where real-time measurement methods provide accurate estimates of delay without information regarding progression quality. This study examines the applicability of different analytical methods in real-time delay estimation procedures by using high-resolution signal and detection data as main input data. Delays were assessed using HCS6, InputOutput, and IQA model and compared with values from microsimulation that served as ground truth.  It was found that HCS6 provides reliable estimates for v/c ratios up to 0.5. InputOutput method in most of the examined scenarios provides reliable delay estimates without information regarding progression quality. When IQA model was enhanced with a procedure for platoon identification, it showed great potential as a method that can provide both, reliable delay estimates with information regarding progression quality for real-time performance assessment. Through the course of study, it was noticed that the proposed Enhanced IQA method lack accuracy in some instances; thus, modifications of the proposed algorithm are recommended in future work.

Adekunle Adebisi, University of Cincinnati
Yi Guo, University of Cincinnati
Bastian Schroeder, Kittelson & Associates, Inc. (KAI)
Jiaqi Ma, University of California, Los Angeles
Burak Cesme, Kittelson & Associates, Inc. (KAI)
Apoorba Bibeka, Kittelson & Associates, Inc. (KAI)
Abby Morgan, Kittelson & Associates, Inc. (KAI)

Connected and Automated Vehicles (CAVs) will potentially improve transportation system performance. Guidance on capacity impact of CAVs at different market penetration rate (MPR) will help agencies to incorporate the effects of CAVs while planning and designing roadways. Traditionally, practitioners use the Highway Capacity Manual (HCM) to assess capacity and evaluate the quality of service for various facilities and systems. While several studies provide insights into the capacity benefits of CAVs, there is a need to quantify CAV effects that can be used to develop HCM guidance. This study estimates the capacity benefits of CAVs at signalized intersections and develops capacity adjustment factors (CAFs) for the HCM. The researchers considered variations in CAV gap/headway settings, platoon lengths, turning movement types (through and left), and left-turn phasing modes (protected versus permitted). Microscopic traffic simulation is used to model CAVs. Results show that performance indicators such as saturation headway gradually improves with increase in CAV MPR, resulting in up to 40% capacity increase at 100% MPR for the protected movements.  For permitted left turns, up to 45% capacity increase could be achieved at 100% MPR. This increase in the permitted left turn capacity is attributed to the vehicle-to-vehicle (V2V) communication that provides advanced information on available gaps in conflicting traffic and reduced follow-up headway time both for the permitted left-turn and the opposing through movement. Based on the capacity results, this paper provides CAF tables for CAVs that can be easily integrated into the HCM and used for planning level guidance by practitioners.

Yonas Emagnu, Nagoya Daigaku
Xin Zhang, Nagoya Daigaku
Miho Iryo-Asano, Nagoya University
Hideki Nakamura, Nagoya University

Vehicle-pedestrian interaction is a significant issue in the study of traffic engineering. Therefore, consideration of crossing pedestrians is one of the critical factors in proper estimation of turning vehicle capacities at signalized intersections. Although individual pedestrian walking speeds and consequent pedestrian platoon maneuver are influenced by traffic signal settings, crosswalk length and bidirectional interactions of pedestrian flows, the existing capacity estimation methods of left-turning vehicles (in the left-hand traffic system)did not fully consider these impacts. This study aims to propose a method to estimate the occupancy time when waiting pedestrian platoon occupies the conflict area, considering the characteristics of pedestrian platoon maneuvers and influencing factors mentioned above. Referring to the existing method to estimate time-dependent pedestrian presence probability distributions, this study modified the parameters of the model using the additional database with higher pedestrian demand conditions. The model revealed that the increase of opposing pedestrian flow significantly reduces spatial variation. Then, the model was applied to propose a method to calculate waiting pedestrian platoon occupancy time and found that it successfully represents the observed occupancy times. Additionally, the comparison of the proposed method occupancy time estimation with the existing pedestrian occupancy time estimation method of HCM revealed that the proposed method can be effectively applied to estimate occupancy time of pedestrians. Furthermore, the result showed that crosswalk length and bi-directional pedestrians flow interaction have significant effect on the distribution of pedestrians’ positions, as a consequence occupancy time will be affected.

Nan Kang, Nanjing Technical University
Xin Zhang, Nagoya Daigaku
Hideki Nakamura, Nagoya University

In recent years, increasing number of roundabouts has been implemented in Japan due to their good performance on both safety and efficiency. To evaluate the roundabout operational performance, entry capacity is one of the most important indices. With the development of vehicular technologies, the appearance of autonomous vehicles (AV) will lead to a mixed flow condition which can cause significant impact on roundabout entry capacity. Thus, this paper aims to examine the influence of AVs on the entry capacity through applying a microscopic traffic simulation. Moreover, in order to clearly classify the impacts of AVs, different aggressive levels and the percentage of AVs involved in flows are considered in this paper. The main findings indicated that aggressive AV can increase entry capacity since shorter gap parameters are defined. Additionally, AVs in entry flow performed greater impact on entry capacity than the condition AVs in circulating flow. Then, based on the findings, an adjustment factor f AV is developed to simplify the estimation process. Through comparing MAPE, it can be concluded that entry capacity under the condition AV mixed flow can be well adjusted by f AV .