Using High-Resolution Controller Data in the Calibration of Traffic Simulation Models
Mosammat Tahnin Tariq, Florida International UniversityShow Abstract
Mohammed Hadi, Florida International University
Rajib Saha, Florida International University
Calibration of traffic simulation models is a critical component of simulation modeling. The increased complexity of the transportation network and the adoption of emerging of vehicle and infrastructure-based technologies and strategies have motivated the development of new methods and data collection to calibrate the simulation models. This study proposes the use of high-resolution controller data, combined with a two-level clustering technique for scenario identifications, and a multi-objective optimization technique for simulation model parameter calibration. The evaluation of the calibration parameters resulting from the multi-objective optimization based on travel time and high resolution controller data measures indicate that the simulation model that uses these optimized parameters produces significantly lower errors in the split utilization ratio, green utilization ratio, arrival on green, and travel time compared to a simulation model that uses the default parameters of the simulation model. When compared with a simulation model that uses calibration parameters obtained based on the optimization of the single objective of minimizing the travel time, the multi-objective optimization solution produces comparably low travel time errors but with significantly lower errors in terms of the high resolution controller data measures. Keywords: Calibration, Microscopic Simulation, High-Resolution Controller Data, Clustering.
Economic-Driven Adaptive Traffic Signal Control
Shan Jiang (email@example.com), Rutgers UniversityShow Abstract
Yufei Huang, Rutgers University
Mohsen Jafari, Rutgers University
Mohammad Jalayer, Rowan University
ABSTRACT With the emerging connected-vehicle technologies and smart roads, the need for intelligent adaptive traffic signal controls is more than ever before. This paper proposes a novel Economic-driven Adaptive Traffic Signal Control (eATSC) model with a hyper control variable – interest rate defined in economics for traffic signal control at signalized intersections. The eATSC uses a continuous compounding function that captures both the total number of vehicles and the accumulated waiting time of each vehicle to compute penalties for different directions. The computed penalties grow with waiting time and is used for signal control decisions. Each intersection is assigned two intelligent agents adjusting interest rate and signal length for different directions according to the traffic patterns, respectively. The problem is formulated as a Markov Decision Process (MDP) problem to reduce congestions, and a two-agent Double Dueling Deep Q Network (DDDQN) is utilized to solve the problem. Under the optimal policy, the agents can select the optimal interest rates and signal time to minimize the likelihood of traffic congestions. To evaluate the superiority of our method, a VISSIM simulation model with classic four-leg signalized intersections is developed. The results indicate that the proposed model is adequately able to maintain healthy traffic flow at the intersection. Keywords : Adaptive Traffic Signal Control, Double Dueling Deep Q Network, Markov Decision Process, Reinforcement Learning
Challenges of Microsimulation Calibration with Traffic Waves using Aggregate Measurements
George Gunter (firstname.lastname@example.org), Vanderbilt UniversityShow Abstract
Sadman Shanto, Texas Tech
Benjamin Seibold, Temple University
Daniel Work, Vanderbilt University
Rabie Ramadan, Temple University
This work explores the challenges associated with calibrating parameters of microscopic models with aggregate speed data, e.g., obtained from roadside sensors. Using the Intelligent Driver Model, we explore how reliably parameters that do not influence the equilibrium flow (i.e., the Fundamental Diagram), but do control the stability of those equilibria, can be determined from aggregate speed data. Using a carefully controlled computational setup, we show that standard loss functions used for calibrating microsimulation models can perform poorly when the true parameters result in an unstable traffic state. Precisely, it is found that all of the considered loss functions frequently return different and incorrect parameter sets that minimize the expected value of the loss function. These results highlight the need for improved loss functions, or even fundamental additions to the model calibration procedure.
A Data-driven Simulation of Naturalistic Driving Environment for Autonomous Vehicle Testing
Xintao Yan, University of MichiganShow Abstract
Shuo Feng (email@example.com), University of Michigan, Ann Arbor
Haowei Sun, University of Michigan
Henry Liu, University of Michigan, Ann Arbor
Microscopic traffic simulation provides a controllable, repeatable, and efficient testing environment for Autonomous Vehicles (AVs). To evaluate AVs' safety performances in the real world, it is critical to test AVs in the simulation of the naturalistic driving environment (NDE). Although human driving behaviors have been extensively investigated in the transportation engineering domain, most existing models were developed for traffic flow analysis purposes. For AV testing purposes, however, the NDE needs to reproduce the critical environment measurements (e.g. speed and spacing) that are distributionally consistent with the real-world driving environment. Only with consistent distributions, simulation results can unbiasedly evaluate AVs' safety performances. To generate such a simulation, we propose a data-driven NDE modeling method, including the initialization method and the vehicle behavioral modeling method. To guarantee the consistency, the NDE is modeled as a Markov chain, and its stationary distribution is twisted as the empirical distributions from the real-world naturalistic driving data, through optimizing the vehicle behavior models. The performance of the proposed method is validated for the highway driving environment by both microscopic and macroscopic measurements. To further validate the simulation capability for AV testing, the generated NDE is utilized to test the safety performance of an AV model. The experiments show that the simulation can effectively evaluate the AV safety performance and produce diverse accident cases that are valuable for the AV development.
COMPARING THE BENEFITS OF CONNECTED AND AUTOMATED VEHICLES TO ONLY AUTOMATED VEHICLES BY CONSIDERING A MULTI-VEHICLE COMMUNICATION SYSTEM
Md Hasibur Rahman, University of Central FloridaShow Abstract
Mohamed Abdel-Aty, University of Central Florida
Yina Wu, University of Central Florida
Connected and automated vehicles (CAVs) are expected to improve both traffic efficiency and safety by reducing the human drivers’ errors. Recently, researchers have focused on the simulation-based approach to evaluate the benefits of CAVs due to the lack of real-world data. However, none of them have attempted to differentiate the benefits of CAVs over automated vehicles (AVs) by incorporating multi-vehicle communication system. This paper aims to fill the existing gap by utilizing separate car-following models for both CAVs and AVs in order to approximate their driving behavior in the Aimsun Next simulation platform. Additionally, a different car-following model is used for the connected vehicles (CVs) without automation by addressing the human driver compliance factor. A well calibrated and validated simulation testbed is developed for the deployment of CAV technologies. To this end, the impacts of CAVs, AVs, and CVs are evaluated based on both traffic efficiency (i.e., travel time) and safety (i.e., traffic conflicts) under various market penetration rates (MPRs). A generalized estimating equation (GEE) model is developed to quantify the travel time improvement for CAVs, AVs, and CVs which suggests that at the same MPR, CAV significantly outperforms AV. For the safety assessment, traffic conflicts are estimated which is further used to develop a Bayesian zero-inflated negative binomial model where results show that CAVs can reduce crash risk more compared to AVs at the same MPR. Also, crash risk analysis based on different vehicle types (CAVs, AVs) shows that CAVs driving behavior is safer compared to the AVs.
A Polynomial Chaos Expansion Based Approach for Efficient And Robust Calibration of Stochastic Transportation Simulation Models
Di Sha, New York UniversityShow Abstract
Kaan Ozbay, New York University
Zilin Bian, New York University
Ding Wang, New York University
The output of a transportation simulation model is stochastic due to the presence of uncertainty in model inputs. Traditionally, Monte Carlo type (MC-type) random sampling is widely used to quantify the uncertainty. However, the low convergence rate makes it prohibitive for complex models in terms of computational burden. This paper proposes an efficient and novel framework for stochastic calibration of transportation simulation models based on polynomial chaos expansion (PCE) which aims to approximate the dependence of simulation model outputs on model inputs by expansion in an orthogonal polynomial basis. The proposed framework contains two parts. The first is to quantify the stochastic simulation outputs due to uncertainties in model inputs. The second is to calibrate model parameters based on the comparison between simulation outputs and observed measures. The proposed framework is applied to calibrate a simulation network developed in SUMO, an open-source microscopic transportation simulation platform. A two-dimensional random space is constructed based on stochastic model inputs. 16 parameters are calibrated using SPSA algorithm. K-S test statistic is used to specify the error in the objective function. Flow and speed are selected as performance measures. Finally, the aggregate error is observed to be reduced from 0.75 to 0.34 after calibration. The flow and speed distributions are also observed to get closer to the observed ones after calibration for each sensor location and each direction. The calibration results reveal that the proposed PCE technique is an efficient way for uncertainty quantification in the process of calibrating a stochastic transportation simulation model.
A Stop Sign Gap Assist Application in a Connected Vehicle Simulation Environment
Mahmoud Arafat, Florida International UniversityShow Abstract
Mohammed Hadi, Florida International University
Thodsapon Hunsanon, Florida International University
Kamar Amine, Florida International University
The assessment of the safety and mobility impacts of Connected Vehicles (CV) and Cooperative Automated Vehicles (CAV) applications is critical to the success of these applications. In many cases, there may be trade-offs in the mobility and safety impacts depending on the setting of the parameters of the applications. This study develops a method to evaluate the safety and mobility benefits of the Stop Sign Gap Assist (SSGA) system, a CV-based application at unsignalized intersections, utilizing a calibrated microscopic simulation tool. The study results confirm that it is critical to calibrate the drivers’ gap acceptance probability distributions in the utilized simulation model to reflect real-world driver behaviors when assessing SSGA impacts. The simulation models with the calibrated gap parameters were then used to assess the impacts of the SSGA. The results showed that SSGA can potentially improve the overall minor approach capacity at unsignalized intersections by approximately 35.5% when the SSGA utilization reaches 100%. However, this increase in capacity depends on the setting of the minimum time gap in the SSGA and there is a clear trade-off between capacity and safety. The analysis indicates that as the minimum gap time used in the SSGA increases, the safety of the intersection increases, showing for example that with the utilization of 8 second-gap at 750 veh/hr/lane main street flow rate, the number of conflicts can decrease by 30% as the SSGA utilization rate increases from 0% to 100%.
Parameter Sensitivity Analysis of A Cooperative Dynamic Bus Lane System with Connected Vehicles
Meng Xie, TUMCREATEShow Abstract
Michael Winsor, TUMCREATE
Tao Ma, Technische Universitat Munchen
Andreas Rau, TUMCREATE
Fritz Busch, Technische Universitat Munchen
Constantinos Antoniou, Technische Universitat Munchen
This paper aims to evaluate the sensitivity of the proposed cooperative dynamic bus lane system with microscopic traffic simulation models. The system creates a flexible bus priority lane that is only activated on demand at an appropriate time with the advanced information and communication technologies, which can maximize the usage of road space. A decentralized multi-lane cooperative algorithm is developed and implemented in a microscopic simulation environment to coordinate lane changing, gap acceptance and car-following driving behavior for the connected vehicles (CVs) in the bus lane and the adjacent lanes. The key parameters for the sensitivity study include the penetration rate and communication range of connected vehicles, considering the transition period and gradual uptake of CVs. Multiple scenarios are developed and compared for analyzing the impact of key parameters on the system’s performance, such as total travel time of all vehicles and travel time variation among buses and private vehicles. The microscopic simulation models showed that the cooperative dynamic bus lane system is significantly sensitive to the variations of the penetration rate and the communication range in a congested traffic state. With a 100% penetration rate, not only the buses but also the private vehicles show improvements in travel time, which represents a win-win situation between private vehicles and public buses. The safety concerns induced by cooperative driving behavior are also discussed in this paper.
MATSim Model Vienna: Analyzing the Socioeconomic Impacts for Different Fleet Sizes and Pricing Schemes of Shared Autonomous Electric Vehicles
Johannes Müller, Austrian Institute of TechnologyShow Abstract
Markus Straub, Austrian Institute of Technology
Asjad Naqvi, International Institute for Applied Systems Analysis
Gerald Richter, Austrian Institute of Technology
Stefanie Peer, Vienna University of Economics and Business
Christian Rudloff, Austrian Institute of Technology
Shared Autonomous Electric Vehicles (SAEVs) are expected to enter the transportation market in the upcoming decades. In this paper, we describe the preparation of a MATSim model for Vienna in which we add this new service as a new transportation mode. We simulate different pricing schemes for various SAEV fleet sizes and analyze their impacts. Our focus is on the impacts in regards of socioeconomic heterogeneity. One main finding of our paper is that the number of SAEV trips does not necessarily decrease for higher fares. It is instead the average travel time of SAEV rides which decreases if the service gets more expensive. Our simulation results for higher pricing schemes show that many people switch from bike or walk mode to SAEV. Public transport is also highly cannibalized by this new service regardless of the price, whereas SAEVs would always replace no more than 10% of car trips. SAEVs help reduce travel times significantly. People who do not have a car available in their household experience the greatest savings in travel time. A similar high share of SAEV trips is done by people older than 35 years. In regards of gender, our results reveal that women tend to use SAEVs for shorter trips.
Cooperative Perception for Estimating and Predicting Microscopic Traffic States to Manage Connected and Automated Traffic
Tao Li, University of California, Los AngelesShow Abstract
Jiaqi Ma (firstname.lastname@example.org), University of California, Los Angeles
Real-time traffic state estimation and prediction are of importance to the traffic management systems. New opportunities are enabled by the emerging sensing and automation technologies to manage connected and automated traffic, particularly in terms of controlling trajectories of automated vehicles. Traffic information from connected and automated vehicles (CAV) and roadside detectors (RSD) have great potential for providing detailed microscopic traffic states (i.e., vehicle speeds, positions). In this paper, we propose a cooperative perception framework for this purpose. The proposed framework based on particle filtering is developed to provide an accurate estimation and prediction of the microscopic states of partially observed traffic systems, while accounting for different sources of errors that intrinsically exist in the system, including those from sensor data, vehicle movement, and process models. Selected freeway and arterial vehicle trajectory datasets from the Next Generation Simulation (NGSIM) program and traffic simulation are applied to test the proposed methodological framework. The accuracy of position and speed estimation is between 50% and 70% when the CAV market penetration rate (MPR) is 12.5%, and between 80% and 90% when the MPR is 50%. The incorporation of RSD data can further increase the accuracy by up to 10% under low CAV MPRs. The framework can also provide an accurate short-term prediction of position and speed with 60% to 90% accuracy. The proposed framework provides efficient and accurate estimations and predictions of detailed microscopic traffic states, creating dynamic traffic environment world models to enable fine control and management of the connected and automated traffic systems.
Integrating Car Following and Lane Changing in a Mixed Traffic Model of Connected Automated Vehicles and Conventional Vehicles
Punyaanek Srisurin, Chulalongkorn UniversityShow Abstract
Alexandra Kondyli, University of Kansas
This study developed mixed-traffic simulation models of Connected Automated Vehicles (CAVs) and Manually-Driven Vehicles (MDVs) at the full-spectrum of penetration rates on a freeway segment by incorporating the car-following and lane-changing models via a linkage to investigate capacity and travel time. The car-following models for CAVs and MDVs were modified from the Full Velocity Difference (FVD) model, while a lane-changing logic was adopted. Stochastic parameters were applied for MDVs to replicate the characteristics of the human drivers, whereas static parameters were adopted to establish the safe decision-making thresholds for CAVs. The CAV algorithm was designed to maintain a sizeable headway between vehicles and milder acceleration for safety and passenger comfort, also equipped with a gap-creation function for enhancing lane-changing maneuvers. The algorithms were coded in JAVA to create a simulation platform, prior to calibrating the model with field data. Eleven mixed-traffic scenarios were simulated, along with parallel simulations in VISSIM, to generate and validate the speed-flow diagrams. The results showed conservative increase in capacities in the range of 25.9% – 26.9%, while travel times decreased by 55.4%, as the CAV penetration rate shifted from 0 to 100 percent. The trajectory analysis indicated that CAVs have an influence on guiding smoother speeds and acceleration rates of MDVs while an MDV is following a CAV. The results suggest that although headways increased with increasing CAV penetration rate, capacity also increased; however, there should be an optimal headway that maximizes the capacity.
Automated Calibration of a Microscopic Traffic Flow Simulation Using Real-World Observations
Michael Harth, Technical University of MunichShow Abstract
Marcel Langer, Audi AG
Klaus Bogenberger, Technische Universitat Munchen
In order to evaluate future mobility services as well as automated driving functions, a virtual environment is required. As a key component of this virtual proving ground, a traffic flow simulation is necessary to represent real-world traffic conditions. Real-world observations, such as historical traffic counts and traffic light state information, provide a basis for the representation of these conditions in the simulation. A main challenge arises in allocating these available real-world data in the virtual world for calibration purposes. In this work, we therefore propose a scalable approach to transfer real-world data, exemplarily taken from the German city Ingolstadt, to a virtual environment for a calibration of a traffic flow simulation. For the allocation of historical real-world data, we use a process which matches real-world measurements with their corresponding locations in the virtual environment. The calibration incorporates the replication of realistic traffic light programs as well as the adjustment of simulated traffic flows. The proposed calibration procedure allows for an automated creation of a calibrated traffic flow simulation of an arbitrary road network given historical real-world observations.
Dynamic Lane Management: One-way Two-lane Freeway with Mixed Traffic
anran li, Beijing University of TechnologyShow Abstract
haijian li, Beijing University of Technology
Xiaohua Zhao, Beijing University of Technology
Intelligent connected vehicles (ICVs) have become the development trend of traffic systems for the advantages of safety and efficiency. But there has been inefficient traffic flow mixed of ICVs and manual vehicles (MVs) during the process of ICVs gradually replacing MVs. In many potential methods to enhance traffic efficiency, researchers had a high expectation of dynamic lane management. However, the researches on the influence of exclusive lane, one of the mainstream dynamic lane management, on mixed traffic flow were not enough. Therefore, the characteristics of mixed traffic flow were studied and compared in this paper with the inﬂuence of diﬀerent dynamic lane management strategies under various traffic densities and ICV penetration rates (p). Four kinds of dynamic lane management strategies (GG, IM, GI, IM) for one-way two-lane freeways were proposed and simulated by an improved cellular automata (CA) model. Through the comparative analysis of traffic capacity, average speed, and congestion rate of each dynamic lane management strategy, it can be concluded that GG is the most effective dynamic lane management strategy under a low p. For a high p, IM is more efficient at low density, while GI is more efficient at high density. Besides, the velocity stability of mixed traffic flow is positively correlated with p. The ﬁndings of this paper may improve the eﬃciency of heterogeneous traﬃc flow mixed with MVs and ICVs.
On the validity of the LWR traffic model
Simone Fagioli, Universita degli Studi dell'AquilaShow Abstract
Adriano Festa, Politecnico di Torino
Corrado Lattanzio, Università degli Studi dell'Aquila
Mario Bellotti, VEM solutions
Giuseppe Cutrupi, VEM solutions
Carmen Criminisi, TIM Telecom Italia
Davide Micheli, TIM Telecom Italia
Giuliano Muratore, TIM Telecom Italia
Aldo Vannelli, TIM Telecom Italia
Empirical model verification and validation is a fundamental step in the study of traffic phenomena. We use a rich geolocalized dataset collected via On-Board-Units systems to tune the parameters of a classic LWR first order macroscopic traffic model and compare it with the empirical data. We discuss the method of processing of the dataset and the calibration technique used to tune the model. Various cases of roads are taken into exam: suburban highways, urban high capacity roads and more likely congested normal streets, where the possible presence of traffic lights and crossings interferes complicates the scenario. An evaluation of the accuracy of the model in the various situations is presented.
A Rigorous Multi-Population Multi-lane Hybrid Traffic Model and its Mean-field Limit for Dissipation of Waves via Autonomous Vehicles.
Nicolas Kardous, University of California, BerkeleyShow Abstract
Amaury Hayat, Rutgers University, Camden
Sean McQuade, Rutgers University, Camden
Xiaoqian Gong, Rutgers University, Camden
Sydney Truong, Rutgers University, Camden
Paige Arnold, Rutgers University, Camden
Alexandre Bayen, University of California, Berkeley
Benedetto Piccoli, Rutgers University, Camden
In this paper, a multi-lane multi-population microscopic model, which presents stop and go waves, is proposed to simulate traffic on a ring-road. Vehicles are divided between human-driven and autonomous vehicles (AV). Control strategies are designed with the ultimate goal of using a small number of AVs (less than 5\% penetration rate) to represent Lagrangian control actuators that can smooth the multilane traffic flow and dissipate the stop-and-go waves. This in turn may reduce fuel consumption and emissions. The lane-changing mechanism is based on three components that we treat as parameters in the model: safety, incentive and cool-down time. The choice of these parameters in the lane-change mechanism is critical to modeling traffic accurately, because different parameter values can lead to drastically different traffic behaviors. In particular, the number of lane-changes and the speed variance are highly affected by the choice of parameters. Despite this modeling issue, when using sufficiently simple and robust controllers for AVs, the stabilization of uniform flow steady-state is effective for any realistic value of the parameters, and ultimately bypasses the observed modeling issue. Our approach is based on accurate and rigorous mathematical models, which allows a limit procedure that is termed, in gas dynamic terminology, mean-field. In simple words, from increasing the human-driven population to infinity, a system of coupled ordinary and partial differential equations are obtained. Moreover, control problems also pass to the limit, allowing the design to be tackled at different scales.
Optimization of Wiedemann-99 Model Parameters for Mixed Traffic Using Vehicular Trajectory Data
Ankit Chaudhari, Indian Institute of Technology, MadrasShow Abstract
Karthik Srinivasan, Indian Institute of Technology, Madras
Bhargava Chilukuri, Indian Institute of Technology, Madras
Martin Treiber, Technische Universitat Dresden
Ostap Okhrin, Technische Universitat Dresden
A new methodology is proposed for calibrating Wiedemann-99 vehicle following parameters for mixed traffic based on trajectory data. The existing acceleration equations of the Wiedemann model are modified to represent more realistic driving behavior. Exploratory analysis of simulation data revealed that different Wiedemann-99 model parameters can lead to similar macroscopic behavior, highlighting the importance of calibration at the microscopic level. Therefore, the proposed methodology is based on optimizing performance measures at the microscopic level (acceleration, speed, and trajectory profiles) to estimate suitable calibration parameters. The performance of the optimized parameters are is validated and compared against other heuristic methods of calibration reported in the literature. It was found that parameters optimized using the proposed methodology performs better than those obtained using heuristic methods. Further, the calibration parameters and goodness of fit with respect tofor the observed data are found to be sensitive to the numerical integration method used to compute the velocity and position of vehicles. The results reveal that the optimized parameter values and consequently, the thresholds that delineate closing, following, emergency braking, and opening regimes vary between two-wheelers and cars. The window (in the relative speed vs. gap plot) for the unconscious following is larger for cars, while the free flow regime is larger for two-wheelers. Under the same stimulus of relative speed and spacing, two-wheelers and cars may be in different regimes and display different acceleration responses. Thus, accurate calibration of parameters for each type of vehicle is essential for developing micro-simulation models for mixed traffic.
Longitudinal Dynamics in Traffic Microsimulation
Shirin Noei (email@example.com), Tennessee Technological UniversityShow Abstract
Xilei Zhao, University of Florida
Simulated accelerations and decelerations are sensitive to maximum acceleration and maximum deceleration inputs, particularly for scenarios with significant truck share. Conventional traffic microsimulation tools estimate maximum acceleration and maximum deceleration using simplified mechanistic models, empirical models, or lookup tables, or some even have maximum acceleration and maximum deceleration as user inputs. Simplified models used in most traffic microsimulation tools may have been a necessary compromise long ago due to software architecture, computational speed limitations, or to minimize data input requirements. Unlike conventional traffic microsimulation tools, several simulation tools (e.g., CarMaker and CarSim) can imitate vehicle performance precisely but cannot or do not even intend to simulate large-scale transportation networks due to computational complexities. Conventional longitudinal controllers rely on constant distance gaps to maximize throughput, constant time gaps to ensure string stability, and constant controller coefficients, potentially reducing throughput or sacrificing safety. This research incorporates driver characteristics, physical, engine, transmission, and drivetrain properties, aerodynamic resistance, rolling resistance, grade resistance, and speed into calculating maximum acceleration, maximum deceleration, minimum safe distance gap, minimum safe time gap, and longitudinal controller coefficients for multiple vehicles at each simulation time step with reasonable accuracy and simulation speed in order to maximize throughput without compromising safety or string stability. Proposed models are verified for fourteen vehicle models operating in autonomous mode over two driving schedules. Results show that maximum acceleration, maximum deceleration, minimum safe distance gap, minimum safe time gap, and longitudinal controller coefficients are sensitive to driver characteristics, vehicle properties, road conditions, and speed.
Gap Acceptance at Blocked Lanes on Urban Two-Way Roads and Evaluation of a Bottleneck Assistant
Sofie Ehrhardt, Karlsruher Institut fur TechnologieShow Abstract
Marvin Baumann, Karlsruhe Institute for Technology (KIT)
H. Sebastian Buck, Karlsruhe Institute of Technology (KIT)
Yang Li, Karlsruhe Institute of Technology (KIT)
Barbara Deml, Karlsruhe Institute of Technology (KIT)
Peter Vortisch, Karlsruhe Institute of Technology (KIT)
The blocking of a lane in one direction on an urban two-lane road (referred to as bottlenecks in this paper), receives little attention in current research, although it is a common scenario in traffic. In order to enable automated driving in this situation in the future, this research gap must be closed. An online study with 86 test persons collects data on gap acceptance at bottlenecks and the user experience and user acceptance of an automated cooperative bottleneck assistant. The participants watched videos created with the microscopic traffic flow simulation software PTV Vissim. They experienced the bottleneck situation from different perspectives and varying oncoming traffic without and with a cooperative bottleneck assistant. The results show that gap acceptance is distributed over a number of time gaps and depends on the individual driver. A time gap of 10 seconds, however, gets accepted in 73.5 % of all cases. The evaluation of the automated assistant shows that the test persons feel less secure with an assistant than they would if they had selected the time gap themselves and expected a passive driving style. A human-machine interface also does not increase the feeling of safety but is nevertheless expected. These results form a basis for the further development of a bottleneck assistant and can be used to calibrate a microscopic traffic flow simulation of such bottlenecks.
DISCLAIMER: All information shared in the TRB Annual Meeting Online Program is subject to change without notice. Changes, if necessary, will be updated in the Online Program and this page is the final authority on schedule information.