Agent-based Traffic Simulation Research Articles

TRAPSim: An agent-based model to estimate personal exposure to non-exhaust road emissions in central Seoul

Non-exhaust emissions (NEEs) from brake and tyre wear cause detrimental health effects, yet their relationship with mobility has not been examined rigorously. We constructed an agent-based traffic simulator to illustrate the coupled problems of emissions, behaviour, and the estimated exposure to PM10 for groups of drivers and subway commuters in Seoul CBD. Having calibrated the parameters, the results regarding the air quality revealed that roughly 25–30% of the roadside PM10 was significantly higher than the background PM10. Additionally, compared to intra-urban cars, pedestrians who commuted for longer periods of time and were exposed to more ambient particles suffered significant health losses; however, drivers only became aware of the health risk when PM10 levels were consistently high for a few days. Compared to the business-as-usual scenario of vehicle entry, a 90% vehicle restriction was able to reduce PM10 by 18–24% and cut the percentage of resident drivers who were at risk. However, it was not effective for subway commuters. Using an agent-based traffic simulator in a health context can provide insights into how exposure and health effects can vary depending on the time of exposure and the form of transportation.

Assessing the impacts of ridesharing services: An agent-based simulation approach

A shift from privately owned vehicles to shared mobility services can affect mobility, energy consumption, and vehicle emissions. Existing literature on ridesharing services has focused on evaluating its traffic and economic impacts. In this study, we propose an integrated framework to analyze the efficiency and environmental benefits of ridesharing on a regional scale. The framework utilizes an agent-based traffic simulation package (i.e., SUMO) to replicate traffic activities for commuting trips in a mid-size city, Chattanooga, Tennessee, based on real-world travel-demand data. We construct scenarios representing different ridesharing strategies and penetrations. The simulation and results analysis show that with a ridesharing ratio of 5%–75% over travel demand in a city scale, many (65%–75%) ridesharing travelers will experience up to a 15-min delay. About 80% of drive-alone travelers will arrive earlier compared with no ridesharing scenario. The average early arrival time would be 5.6 min for all drive-alone travelers. The results also show ridesharing services can achieve a 2%–50% reduction in total city-scale vehicle emissions and energy consumption compared with the no ridesharing scenario. The framework and results of this study can be helpful to transportation practitioners to evaluate environmental benefits when implementing ridesharing services on a city scale.

RISK ANALYSIS OF CYCLING ACCIDENTS USING A TRAFFIC DEMAND MODEL

Abstract. In Germany, accidents are collected nationwide, including also bicycle accidents. To determine accident-prone locations, it is necessary to not only look at the number of accidents but also in relation to the absolute number of cyclists traversing that location. Thus, this study exploits a collection of bicycle accidents in combination with estimated cyclist volumes on street level in Hanover (Germany). The basis for the generated bicycle volumes is the resulting origin-destination demand for bicycle mode from an agent-based traffic simulation model. A normalization of the accidents by an absolute bicycle volume allows to estimate a risk score and to compare high frequented ways with less popular minor paths in an objective manner. This method is used to show locations with comparatively high risk for cyclists. Besides highlighting these spots on a map, e.g. for city planners, the resulting risk scores can be integrated into bicycle routing to avoid those areas for future trips.

Adaptivity in multi-level traffic simulation using experimental frames

Balancing between simulation performance and simulation accuracy is difficult. Most simulators focus on optimizing only one of these two aspects. Which comes at the cost of the other. In this work we demonstrate a framework to dynamically balance between computational performance and simulation accuracy based on the context of the simulation. This allows to maximize both performance and accuracy when possible during the execution of the simulation. To do this, we present a method to apply adaptive abstraction in a large-scale agent-based traffic simulation. This method uses the concept of experimental frames for simulation models in order to keep track of model validity given the current simulation context. Furthermore, we present a custom developed state-of-the-art multi-level traffic simulator that includes adaptive abstraction in the core of its simulation architecture. We validate the proposed methods in a realistic multi-level traffic simulation use case executed in an urban environment.

Impact of electric vehicle charging – An agent‐based approach

Several European countries have launched programs to increase the market penetration of battery electric vehicles (BEV). Similarly, politicians in Switzerland have targeted a 15% BEV share of new car registrations by 2022. As each electric car increases the power demand, new challenges are posed to the operation of existing distribution grid infrastructure. Here, a new bottom-up physical approach is presented that couples agent-based traffic simulations through an unsteady vehicle power consumption model with distribution grid power flow simulations. The impacts on hourly powerline loads from charging a car fleet with an 8.5% BEV share are quantified in the real distribution grid for the canton of Zurich. The grid is composed of 12,000 buses and 9,800 powerlines, providing power to 398,000 individual customers. Results indicate that the risk of overloaded powerlines is highest in low-level distribution grids. In our most critical future scenario, with simultaneous 8.5% BEV charging at 8 pm with 11 kW, peak line loads reach up to 132% of rated capacity. Hence, in a potential energy transition towards a decarbonised future, each individual distribution grid could face critical loads at specific temporal and spatial bottlenecks. Thus, grids should be assessed individually to limit uncertainties and risks of critical power system situations.

Impacts of Automated Mobility-on-Demand on traffic dynamics, energy and emissions: A case study of Singapore

Technological advancements have focused increasing attention on Automated Mobility-on-Demand (AMOD) as a promising solution that may improve future urban mobility. During the last decade, extensive research has been conducted on the design and evaluation of AMOD systems using simulation models. This paper adds to this growing body of literature by investigating the network impacts of AMOD through high-fidelity activity- and agent-based traffic simulation, including detailed models of AMOD fleet operations. Through scenario simulations of the entire island of Singapore, we explore network traffic dynamics by employing the concept of the Macroscopic Fundamental Diagram (MFD). Taking into account the spatial variability of density, we are able to capture the hysteresis loops, which inevitably form in a network of this size. Model estimation results at both the vehicle and passenger flow level are documented. Environmental impacts including energy and emissions are also discussed. Findings from the case study of Singapore suggest that the introduction of AMOD may bring about significant impacts on network performance in terms of increased VKT, additional travel delay and energy consumption, while reducing vehicle emissions, with respect to the baseline. Despite the increase in network congestion, production of passenger flows remains relatively unchanged.

AVDM: A hierarchical command-and-control system architecture for cooperative autonomous vehicles in highways scenario using microscopic simulations

Microscopic agent-based traffic simulation is an important tool for the efficient and safe resolution of various traffic challenges accompanying the introduction of autonomous vehicles on the roads. Both the variety of questions that can be asked and the quality of answers provided by simulations, however, depend on the underlying models. In mixed traffic, the two most critical models are the models describing the driving behaviour of humans and AVs, respectively. This paper presents AVDM (Autonomous Vehicle Driving Model), a hierarchical AV behaviour model that allows the holistic evaluation of autonomous and mixed traffic by unifying a wide spectrum of AV functionality, including long-term planning, path planning, complex platooning manoeuvres, and low-level longitudinal and lateral control. The model consists of hierarchically layered modules bidirectionally connected by messages and commands. On top, a high-level planning module makes decisions whether to join/form platoons and how to follow the vehicle’s route. A platooning manoeuvres layer guides involved AVs through the manoeuvres chosen to be executed, assisted by the trajectory planning layer, which, after finding viable paths through complex traffic conditions, sends simple commands to the low-level control layer to execute those paths. The model has been implemented in the BEHAVE mixed traffic simulation tool and achieved a 92% success rate for platoon joining manoeuvres in mixed traffic conditions. As a proof of concept, we conducted a mixed traffic simulation study showing that enabling platooning on a highway scenario shifts the velocity-density curve upwards despite the additional lane changing and manoeuvring it induces.

Traffic life: temporal dynamics and regulatory dimensions in agent-based transport simulations

ABSTRACT The article discusses the interplay of normative and temporal dynamics in agent-based traffic simulations (ABM). From a media-historical perspective, it focuses on the TRANSIMS simulation system as a seminal example for an ABM ‘mindset’ in transportation and infrastructure simulation. ABM explicitly links traffic simulation with the broader focus of mobility studies by connecting the mere physicality of transport dynamics with a sociality of agents claiming to be descriptive of real-life structures. First, TRANSIMS elevates the examination of traffic dynamics to a meta-level of urban infrastructure design where individual timing and purposeful agent behaviors are placed at the heart of traffic systems. Second, TRANSIMS provides a ‘virtual testbed’ by generating traffic scenarios which eventually lead to situations that meet certain normative limits or regulatory guidelines. And third, ABM often display a strong tendency towards a methodological individualism which requires a ‘theory guidance’ by disciplines like sociology, media theory, or political science to challenge the oftentimes oversimplifying parameters of their ‘artificial sociality’. Consequently, with regard to the scope of mobility studies, ABM can be understood as a medium which negotiates conceptual and interdisciplinary differences and thereby transcends the solely pragmatist notion of ‘virtual testbeds’ as unmitigated optimization tools.

Multi-objective eco-routing for dynamic control of connected & automated vehicles

The advent of intelligent vehicles that can communicate with infrastructure as well as automate the movement provides a range of new options to address key urban traffic issues such as congestion and pollution, without the need for centralized traffic control. Furthermore, the advances in the information, communication, and sensing technologies have provided access to real-time traffic and emission data. Leveraging these advancements, a dynamic multi-objective eco-routing strategy for connected & automated vehicles (CAVs) is proposed and implemented in a distributed traffic management system. It is applied to the road network of downtown Toronto in an in-house agent-based traffic simulation platform. The performance of the proposed system is compared to various single-objective optimizations. Simulation results show the significance of incorporating real-time emission and traffic state into the dynamic routing, along with considering the expected delays at the downstream intersections. The proposed multi-objective eco-routing has the potential of reducing GHG and NOx emissions by 43% and 18.58%, respectively, while reducing average travel time by 40%.

An Agent-Based Simulation Approach for Evaluating the Performance of On-Demand Bus Services

On-demand multi-passenger shared transport options are increasingly being promoted as an influential strategy to reduce traffic congestion and emissions and improve the convenience and travel experience for passengers. These services, often referred to as on-demand public transport, are aimed at meeting personal travel demands through the use of shared vehicles that run on flexible routes using advanced tools for dynamic scheduling. This paper presents an agent-based traffic simulation model that was developed to evaluate the performance of on-demand public transport and compare it with existing scheduled bus services using a case study of the inner city of Melbourne in Australia. The key performance measures used in the comparative evaluation included quality of service and passenger experience in terms of waiting times, the efficiency of service and operations in terms of hourly vehicle utilization, and system efficiency in terms of trip completion rates, passenger kilometers travelled and total passenger trip times. The results showed significant benefits for passengers who use on-demand bus services compared to scheduled bus services. The on-demand bus service was found to reduce average total passenger waiting times by 89% during the Morning Peak; by 78% during the Mid-Day period; by 81% during the Afternoon Peak; and by more than 95% during other periods of the day. From an operator’s perspective, the on-demand services were found to achieve around 70% vehicle utilization rates during peak hours compared to a utilization rate not exceeding 16% for the scheduled bus services. Even during off-peak periods, the occupancies for on-demand services were almost twice the vehicle occupancies for scheduled bus services. In terms of system efficiency, the on-demand services achieved a trip completion rate of 85% compared to a trip completion rate of 67% for the scheduled bus services. The total passenger-kilometers travelled was similar for both scheduled and on-demand bus services, which refutes claims that on-demand bus services induce more kilometers of travel. The trip completion times were around 55% shorter for on-demand bus services compared to scheduled services, which represents a significant saving in travel time for users. Finally, the paper presents average emissions per completed trip for both types of services and shows a significant reduction in emissions for on-demand services compared to conventional bus services. These include, on average, a 48% reduction in CO2 emissions per trip; 82% reduction in NO emissions per trip; and 41% reduction in p.m.10 emissions per trip. These findings clearly demonstrate the superior benefits of on-demand bus services compared to scheduled bus services.

Evaluating the impacts of shared automated mobility on-demand services: an activity-based accessibility approach

Autonomous vehicle (AV) technologies are under constant improvement with pilot programs now underway in several urban areas worldwide. Modeling and field-testing efforts are demonstrating that shared mobility coupled with AV technology for automated mobility on-demand (AMoD) service may significantly impact levels of service and environmental outcomes in future cities. Given these rapidly emerging developments, there is an urgent need for methods to adequately quantify the economic impacts of new vehicle technologies and future urban mobility policy. In this paper, we show how broader user-centric impacts can be captured by the activity-based accessibility (ABA) measure, which takes advantage of the rich data and outcomes of utility-maximization activity-based models and its interaction with mesoscale agent-based traffic simulation frameworks. Using the SimMobility simulator, we evaluate shared AMoD strategies applied to a Singapore micromodel city testbed. A near-future strategy of exclusive availability of AMoD service in the central business district (CBD), and a further-horizon strategy of the full operation of AMoD city-wide in the absence of other on-demand services, were tested and evaluated. Our results provide insights into the income and accessibility effects on the population under the implementation of shared and automated mobility policies. The outcomes indicate that the city-wide deployment of AMoD results in greater accessibility and network performance. Moreover, the accessibility of low-income individuals is improved relative to that of mid- and high-income individuals. The restriction of AMoD to the CBD along with the operation of other on-demand services, however, provides a certain level of disbenefit to segments of the population in two exceptional cases. The first is to high-income individuals who live in a suburban zone and rely heavily on on-demand services; the second is to mid-income residents that have excellent public transportation coverage with close proximity to the CBD. We further establish the efficacy of the ABA measure, as these findings motivate the need for measuring socioeconomic impacts at the individual level. The work presented here serves as a foundation for policy evaluation in real-world urban models for future mobility paradigms.

Smart technology implementation for road traffic management

The escalation of road traffic appears to be a tremendous problem. Various metropolises are influenced by traffic flow congestion and the growth of emissions from petrol usage. In big agglomerations, the expanding quantity of private cars and public transport has caused traffic problems. They have a harmful effect on economy, ecosystem, and on the quality of life in general. It is vital to obtain smart solutions for road traffic management. In this paper, authors propose a way to solve this problem by using smart traffic regulation, which is a part of the bigger smart logistics concept. Agent-based traffic simulation has been chosen to perform this research. This type of modeling is related to the object-oriented way of coding. For modeling and experimental simulation of the intersection in St. Petersburg, AnyLogicmodeling software was used. The results show that proposed algorithm allowed to reduce the average waiting time by 37%. Moreover, the average waiting car number at the intersection has been dropped by 2.5 times after applying the new solution. Thus, projected way of reducing road overload on the selected intersection in St. Petersburg displayed excellent outcomes. However, implementation of the algorithm on other infrastructure objects requires further investigation and analysis.

Consistent Population Synthesis With Multi-Social Relationships Based on Tensor Decomposition

Social relationships have a strong influence on individual travel behavior and, consequently, on travel demand. However, most current literatures on population synthesis, which is the fundamental building block of disaggregated travel demand forecasting and agent-based traffic simulation, only considers the household impact. This paper makes two contributions in this regard. First, a methodological issue is identified: the existence of multiple social relationships (e.g., a dual set of constraints from social institutions or structures) makes it more difficult to generate a consistent synthetic population, meaning that this population satisfies constraints from more than one type of social organizations. A tensor decomposition method is then proposed to generate a consistent population with multi-social relationships. To our knowledge, this is the first time that this type of methodological issue has been addressed. Our sample-based method constitutes an improvement compared to existing approaches in that it can respect constraints from multiple social organizations without reducing accuracy. A numerical test concerning individual, household, and enterprise, using Chinese national population and economic census data, indicates that the new method can lead to stable and relatively small errors in total. The source code is available from https://github.com/PeijunYe/MulSocPopSyn.git .

Driver profiling – Data-based identification of driver behavior dimensions and affecting driver characteristics for multi-agent traffic simulation

This paper focusses on the role of driver individuality in the field of cognitive driver behavior modeling for the prospective safety impact assessment of advanced driver assistance systems (ADAS) and automated driving functions. Virtual traffic simulation requires valid models for the environment, the vehicle and the driver. Especially modeling human driver behavior is a major challenge, which in recent years has already led to the development of various driver models for the purpose of virtual simulation. Modeling human behavior in traffic with a precise representation of human cognition, capability and individuality, are crucial demands, which require thorough investigation and understanding of the human driver. Current driver behavior models often leave aside the aspect of driver individuality and lack the consideration of differences in driving behavior between different drivers. To take into account all the aspects from complex human cognitive processes to individual differences in action implementation, the Stochastic Cognitive Model (SCM) was developed. The SCM is based on five subcomponents: gaze control, information acquisition, mental model, action manager and situation manager (=decision making process) and action implementation. The aim of the present study is to provide a basis for establishing a solid logic for the integration of driver individuality into the current structure of the SCM by creating a new submodule that takes into account several behavior affecting driver characteristics. This subcomponent controls the stochastic variance in several driver behavior parameters, such as velocity or comfort longitudinal acceleration. In a representative driving simulator study with 43 participants, driver behavior on the highway was investigated and thoroughly analyzed. Information about several relevant driver characteristics and personality traits of the participants was collected and a logical hierarchical model was set up to cluster several dependent and independent variables into four layers: independent manifest driver variables, such as age or gender (Level 1), latent driver personality factors, such as thrill seeking or anxiety (Level 2), driver behavior dimensions, such as dynamics and law conformity (Level 3), and various dependent driver behavior parameters, such as velocity, acceleration or speed limit violation (Level 4). Multiple linear regression analyses were run to find the individual driver characteristics and personality traits, by which most of the stochastic variance in the measured driver behavior parameters can be explained. Subsequently, a principal component analysis (PCA) was run to test, if the previously clustered driver behavior parameters were loading on the presumed behavioral dimensions on the third level of the model to identify significant components of driver behavior, such as dynamics or law conformity. Results of the present study show significant correlations between driver characteristics and driver behavior parameters. According to the results of the PCA, variability in driver behavior can be explained to a great extent by three largely independent components: (1) Speed and cruise control, (2) Dynamics and (3) Driver performance. With the consideration of driver individuality in driver behavior models for the agent-based traffic simulation, validity of the results from prospective safety impact assessment analyses of automated driving functions can be enhanced. Beyond that, the findings of the current study can be used as a solid basis for the development of adaptive functions in the field of vehicle automation, considering the different driving skills and preferences of drivers with different individual profiles.

Effects of scaling down the population for agent-based traffic simulations

Agent-based transport models simulate the travel demand of individuals at a fine resolution. However, when large road networks and numerous agents are simulated, they require long runtimes. In order to achieve reasonable runtimes, a common strategy is to randomly subsample the entire population. The capacity of the road network is scaled down proportionally. Previous studies have found that scaling affects the simulated agents’ travel time and distance, although the reasons behind that and the size of the impact remain unclear for the scientific community. In this paper, we present a systematic analysis of the consequences of scaling down the synthetic population. With this goal, we simulate a scenario with different scale factors between 1 and 100% using the multi-agent simulator MATSim. Different numbers of model iterations and two road network resolutions are compared. The output of the simulations shows how the scaling process modifies the model runtime, the travel times or the link volumes.

Distributed Agent-Based Traffic Simulations

Modeling and simulation play an important role in transportation networks analysis. With the widespread of personalized real-time information sources, it is relevant for the simulation model to be individual-centered. The agent-based simulation is the most promising paradigm in this context. However, representing the movements of realistic numbers of travelers within reasonable execution times requires significant computational resources. It also requires relevant methods, architectures and algorithms that respect the characteristics of transportation networks. In this paper, we tackle the problem of using high-performance computing for agent-based traffic simulations. To do so, we define two generic agent-based simulation models, representing the existing sequential agent-based traffic simulations. The first model is macroscopic, in which travelers do not interact directly and use a fundamental diagram of traffic flow to continuously compute their speeds. The second model is microscopic, in which travelers interact with their neighbors to adapt their speeds to their surrounding environment. We define patterns to distribute these simulations in a high-performance environment. The first distributes agents equally between available computation units. The second pattern splits the environment over the different units. We finally propose a diffusive method to dynamically balance the load between units during execution. The results show that agent-based distribution is more efficient with macroscopic simulations, with a speedup of 6 compared to the sequential version, while environmentbased distribution is more efficient with microscopic simulations, with a speedup of 14. Our diffusive load-balancing algorithm improves further the performance of the environment based approach by 150%.

Comparison between ad-hoc demand responsive and conventional transit: a simulation study

Considering the sprawl of cities, conventional public transport with fixed route and fixed schedule becomes less efficient and desirable every day. However, emerging technologies in computation and communication are facilitating more adaptive types of public transport systems, such as demand responsive transport that operates according to real-time demand. It is crucial to study the feasibility and advantages of these novel systems before implementation to prevent failure and financial loss. In this work, an extensive comparison of demand responsive transport and conventional public transport is provided by incorporating a dynamic routing algorithm into an agent-based traffic simulation. The results show that replacing conventional public transport with demand responsive transport will improve the mobility by decreasing the perceived travel time by passengers without any extra cost under certain circumstances. The simulation results are confirmed for different forms of networks, including a real-world network proving the potential of demand responsive transport to solve the challenge of underutilised conventional public transport in suburban areas with low transport demand.

Adaptive traffic signal control with actor-critic methods in a real-world traffic network with different traffic disruption events

The transportation demand is rapidly growing in metropolises, resulting in chronic traffic congestions in dense downtown areas. Adaptive traffic signal control as the principle part of intelligent transportation systems has a primary role to effectively reduce traffic congestion by making a real-time adaptation in response to the changing traffic network dynamics. Reinforcement learning (RL) is an effective approach in machine learning that has been applied for designing adaptive traffic signal controllers. One of the most efficient and robust type of RL algorithms are continuous state actor-critic algorithms that have the advantage of fast learning and the ability to generalize to new and unseen traffic conditions. These algorithms are utilized in this paper to design adaptive traffic signal controllers called actor-critic adaptive traffic signal controllers (A-CATs controllers).The contribution of the present work rests on the integration of three threads: (a) showing performance comparisons of both discrete and continuous A-CATs controllers in a traffic network with recurring congestion (24-h traffic demand) in the upper downtown core of Tehran city, (b) analyzing the effects of different traffic disruptions including opportunistic pedestrians crossing, parking lane, non-recurring congestion, and different levels of sensor noise on the performance of A-CATS controllers, and (c) comparing the performance of different function approximators (tile coding and radial basis function) on the learning of A-CATs controllers. To this end, first an agent-based traffic simulation of the study area is carried out. Then six different scenarios are conducted to find the best A-CATs controller that is robust enough against different traffic disruptions. We observe that the A-CATs controller based on radial basis function networks (RBF (5)) outperforms others. This controller is benchmarked against controllers of discrete state Q-learning, Bayesian Q-learning, fixed time and actuated controllers; and the results reveal that it consistently outperforms them.

Lane Selection Behavior Modeling in an Agent-Based Traffic Simulation

Lane Selection Behavior Modeling in an Agent-Based Traffic Simulation

Relaxing Synchronization in Parallel Agent-Based Road Traffic Simulation

Large-scale agent-based traffic simulation is computationally intensive. Parallel computing can help to speed up agent-based traffic simulation. Parallelization of agent-based traffic simulations is generally achieved by decomposing the road network into subregions. The agents in each subregion are executed by a Logical Process (LP). There are data dependencies between LPs which require synchronization of LPs. An asynchronous protocol allows LPs to progress and communicate asynchronously. LPs use lookahead to indicate the time to synchronize with other LPs. Larger lookahead means less frequent synchronization operations. High synchronization overhead is still a major performance issue of large-scale parallel agent-based traffic simulations. In this article, two methods to increase the lookahead of LPs for an asynchronous protocol are developed. They take advantage of uncertainties in traffic simulation to relax synchronization without altering simulation results statistically. Efficiency of the proposed methods is investigated in the parallel agent-based traffic simulator SEMSim Traffic. Experiment results showed that the proposed methods are able to reduce overall running time of the parallel simulation compared to existing methods.