Multi-agent Transport Simulation Research Articles

Autonomous Vehicles and their Impact on Parking Search

Parking is a major constraint for car users and therefore an important factor in mode choice decisions. In this paper we introduce a model to simulate parking search behavior for cars within a multiagent transport simulation, including full simulation of all steps of parking search, such as walking to and from the vehicle. This is combined with the capabilities of privately owned autonomous vehicles (AVs), which may park automatically, often in other locations than conventional cars, once they are not in use. Three different strategies for AVs to park are developed: (1) Conventional parking search, (2) parking at a designated AV lot, and (3) empty cruising, where vehicles do not use any parking space, but keep on driving. We apply the simulation model to a residential neighborhood in central Berlin, where parking pressure is generally high and apply different shares of AV usage to the synthetic population used. This allows a detailed evaluation of effects for both AV and conventional vehicle owners. Results suggest that the usage of designated parking lots may be the most beneficial solution for most users, with both vehicle wait times and parking search durations being the lowest.

Comparing Parking Strategies of Autonomous Transit On Demand with Varying Transport Demand

Autonomous transit on demand are increasingly considered to become a viable substitute for taxi services. AVs can be managed through a centralized controlling system, targeting system optimization rather than user optimality. This centralized control can enable a more efficient, strictly-adhered-to parking strategy to reduce inefficient empty traveling. In this project, four different parking strategies are implemented in the AV extension of MATSim (Multi-agent transport simulation), namely demand-based roaming, parking on the street, parking in depots and a mixed strategy of parking on the street and in depots. The influence of different PT demand levels on the different parking strategies was explored, showing that the shared system is robust to varying levels of demand, and that the different parking strategies trade off user convenience for operational cost. The road parking strategy appears to be the best for consolidating rides into larger vehicles, especially for the increased demand scenario.

CONGESTION EFFECTS OF AUTONOMOUS TAXI FLEETS

Fleets of shared Autonomous Vehicles (AVs) could replace private cars by providing a taxi-like service but at a cost similar to driving a private car. On the one hand, large Autonomous Taxi (AT) fleets may result in increased road capacity and lower demand for parking spaces. On the other hand, an increase in vehicle trips is very likely, as travelling becomes more convenient and affordable, and additionally, ATs need to drive unoccupied between requests. This study evaluates the impact of a city-wide introduction of ATs on traffic congestion. The analysis is based on a multi-agent transport simulation (MATSim) of Berlin (Germany) and the neighbouring Brandenburg area. The central focus is on precise simulation of both real-time AT operation and mixed autonomous/conventional vehicle traffic flow. Different ratios of replacing private car trips with AT trips are used to estimate the possible effects at different stages of introducing such services. The obtained results suggest that large fleets operating in cities may have a positive effect on traffic if road capacity increases according to current predictions. ATs will practically eliminate traffic congestion, even in the city centre, despite the increase in traffic volume. However, given no flow capacity improvement, such services cannot be introduced on a large scale, since the induced additional traffic volume will intensify today’s congestion.

Vehicle routing algorithms based on a route reservation approach

Traffic congestion remains a serious problem in transportation networks. Widely used navigation systems can only react to the presence of traffic jams but not to prevent their creation. One of the possibilities to prevent congestion is to manage road traffic within the urban area. This work considers a route reservation approach with possibility to reroute a vehicle during a journey. This approach decomposes road segments into time-spatial slots and for every vehicle it makes the slots reservation for the corresponding route. Since the travel time in real networks cannot be determined precisely and can be considered as stochastic, we propose to use a rerouting procedure to minimize the traveling time. The experiments are carried out in microscopic simulation of a real-world traffic environment in the transportation network of Samara, Russia, using multi-agent transport simulation MATSim.

Agent-Based Modelling of Locating Public Transport Facilities for Conventional and Electric Vehicles

This paper proposes an agent-based transport facility development model for both Conventional Vehicles (CVs) and Electric Vehicles (EVs), as a key component of an agent-based land use-transport model, SelfSim. The model attempts to simultaneously locate public parking lots, refuelling stations, charging stations and charging posts at parking lots with the consideration of competitions and interactions between the facilities. The facility development model is composed of a link-based model and node-based model that are used to simulate the development of link-based (e.g., replenishing stations) and node-based facilities (e.g., parking lots), respectively, based on the spatial and temporal disaggregate demand. The demand is extracted from the activity-based simulation with MATSim-EV that is an EV extension of MATSim (Multi-Agent Transport Simulation). In the model, facility agents are defined with several specific attributes and behavioural rules, and act the role of locating transport facilities to accommodate the demand. Finally, both global and local sensitivity analyses are applied to fully test the model in several experiments set up based on a Chinese medium-sized city, Baoding. The global SA that is based on Elementary Effect Method is firstly applied to quantify the extent to which the twelve model outputs of interest are sensitive to forty key model parameters, resulting in nine significantly important parameters; Then the Once-At-A-Time (OAT)-based local SA is used to provide further insight into how these important parameters influence the model outputs of interest over years. The SA results are expected to be useful for model calibration, and how the SA results can be used to calibrate the model is discussed.

Output variability caused by random seeds in a multi-agent transport simulation model

Dynamic transport simulators are intended to support decision makers in transport-related issues, and as such it is valuable that the random variability of their outputs is as small as possible. In this study we analyse the output variability caused by random seeds of a multi-agent transport simulator (MATSim) when applied to a case study of Santiago de Chile. Results based on 100 different random seeds shows that the relative accuracies of estimated link loads tend to increase with link load, but that relative errors of up to 10 % do occur even for links with large volumes. Although the proportion of links having large relative errors is roughly the same for all of the investigated seeds, it is shown the variations of individual link loads between seeds largely dominate variations between the two last iterations within a seed.

Multi-agent urban transport simulations using OD matrices from mobile phone data

Although new available big data sources have revealed themselves to be extraordinarily useful for transport demand modelling, they have not come into widespread use due to the justifiable privacy concerns of data stewards. In this study, we step back and re-evaluate the way in which mobile phone telco data can be introduced for the task of transport and land-use policy evaluation, travel demand forecasting and transport infrastructure testing through large-scale transportation simulations. We investigated that question by deploying a multi-agent transport simulation driven primarily by hourly-aggregated telco Origin-Destination (OD) matrices. We address the principal four challenges: spatial and temporal disaggregation, mode imputation and route choice. For temporal disaggregation, we propose a convolution with an exponential kernel method. As for transport mode imputation, a supervised-learning framework is designed. The simulation results are compared against traffic count data and public transport smart card transactions, showing accurate patterns for private cars but overestimated public transport demand in the morning peak. Lastly, we set the future steps for the improvement of simulations driven by aggregated mobile phone data.

The Dynamic Optimization of the Departure Times of Metro Users during Rush Hour in an Agent-Based Simulation: A Case Study in Shenzhen, China

As serious traffic problems have increased throughout the world, various types of studies, especially traffic simulations, have been conducted to investigate this issue. Activity-based traffic simulation models, such as MATSim (Multi-Agent Transport Simulation), are intended to identify optimal combinations of activities in time and space. It is also necessary to examine commuting-based traffic simulations. Such simulations focus on optimizing travel times by adjusting departure times, travel modes or travel routes to present travel suggestions to the public. This paper examines the optimal departure times of metro users during rush hour using a newly developed simulation tool. A strategy for identifying relatively optimal departure times is identified. This study examines 103,637 person agents (passengers) in Shenzhen, China, and reports their average departure time, travel time and travel utility, as well as the numbers of person agents who are late and miss metro trips in every iteration. The results demonstrate that as the number of iterations increases, the average travel time of these person agents decreases by approximately 4 min. Moreover, the latest average departure time with no risk of being late when going to work is approximately 8:04, and the earliest average departure time with no risk of missing metro trips when getting off work is approximately 17:50.

Sensitivity analysis of integrated activity-based model: using MATSim as an example

This paper attempts to apply both global and local sensitivity analyses (SA) to fully test the integrated activity-based model (ABM) within several experiments based on a Chinese medium-sized city, Baoding. MATSim (Multi-agent Transport Simulation) that is a typical integrated ABM is used as an example for the SA. The global SA, which is based on the elementary effect method, is firstly applied to identify the influential parameters. Then the once-at-a-time (OAT)-based Local SA is employed to further reveal the relationship between the influential parameters and the outputs of interest, such as traffic flow. The SA results show the extent to which and how three influential MATSim parameters (population scaling factor, the number of iterations and time step size) influence the outputs of interest. In addition, the SA results of MATSim suggest that the parameters of time mutation rate and performing utility can heavily influence the outputs of interest and properly setting them can optimize the daily plans of agents. Finally, this paper concludes with suggestions on how to wisely use the SA findings for both MATSim and other ABM users.

Integrating explicit parking search into a transport simulation

Abstract: Explicit parking search is not widely integrated into transport simulation and transport models. In this paper, the integration of a parking search simulation into MATSim (Multi-Agent Transport Simulation) is demonstrated. This includes the integration into the agent's simulation logic using within-day re-planning methodology, a separation up of car trips into several segments for each stage of the trip, a parking search behavior and a data structure for parking infrastructure. The parking search model is applied in a case study for an area in Berlin, Germany. Compared to a standard simulation without parking search, results suggests that parking search traffic sums up to 20 per cent of the overall traffic in a residential area and has a significant impact on the overall travel times of agents traveling by car.

An evacuation route choice model based on multi-agent simulation in order to prepare Tsunami disasters

ABSTRACTMinimising the damage caused by natural disasters such as tsunamis in vulnerable coastal areas requires the development of an emergency action plan. An evacuation plan can reduce the damaging impact of an unexpected tsunami. One of the most important factors in developing an action plan is a guideline for choosing proper evacuation routes in order to minimise loss of life and property. In this study, an evacuation network is designed to develop an evacuation plan based on a hierarchical network design problem used in communication networks, public transport, and social organisation network designs. Experimental analyses have been carried out in order to find effective evacuation plans at Haeundae Beach, Busan, Korea, using multi-agent transportation simulation (MATSim). In order to design an effective hierarchical evacuation network, a hierarchical evacuation network design is developed using the concepts of hub location, clustering, and network design. Further, we consider contraflow in the evacuation network, as well as determination of corridor lines by the minimum distance rule, maximal flow rule, maximum accessible rule, and minimum travel time rule. As a result of running each scenario using MATSim, a criterion for establishing the corridor line is evaluated. An optimal evacuation route is found to minimise overcrowded queues of evacuees, when the total evacuation time from the area of the tsunami is approximately 3 hours and 30 minutes.

Multi-agent transport simulations and economic evaluation

Tolls are frequently discussed policies to reduce traffic in cities. However, road pricing measures are seldom implemented due to high investments and unpopularity. Transportation planning tools can support planning authorities by solving those problems if they take into account the following aspects: – Demographic attributes like income and time constraints – Time reactions to the policy – Schedule changes of population’s individuals during the whole day Our approach uses multi-agent simulations to model and simulate full daily plans. Each of our agents has a utility function that appraises the performance of a typical, microscopically simulated day. The sum of all utility changes to a policy change can be interpreted as the change in the system’s welfare thus the economic evaluation of a measure straightforward. The approach is tested with travel behavior of the Zurich metropolitan region in Switzerland. Several tolling schemes are investigated. It is shown that the simulation can be used to model travelers’ reactions to time-dependent tolls in a way most existing transportation planning tools are not able to do. It is demonstrated that route adjustment only, as is done in many traditional transport planning packages, results in no economic gains from the tolls. As time-dependent tolls are a much-debated subject in transportation politics, the ability to fully model such tolls and the reactions of travelers may help to find better toll schemes. In a world where individuals have more and more freedom to schedule their daily plans, agent-based simulations offer an intuitive way to research complex topics with lots of interdependencies.

Towards an Agent-based, Integrated Land-use Transport Modeling System

This paper reports on initial steps of an integration of the microscopic land-use simulation system SILO (Simple Integrated Land- Use Orchestrator) and the agent-based transport simulation system MATSim (Multi-Agent Transport Simulation). It is shown how information can be transferred from the land-use model to the transport model in an agent-oriented fashion and how MATSim can be used as a transport model within the SILO framework in lieu of an aggregate transport model, which SILO has been coupled with up to now. It is shown that results of the previous model structure can be reproduced by the new fully microscopic modeling system based on SILO and MATSim. It is discussed how an agent-based transfer of information can also be established in the reverse direction, i.e. from the transport model to the land-use model based on the implementation of a query architecture. Finally, it is discussed how an integration of SILO and MATSim can help addressing additional current demands for research.

Keynote II Computational and Software Engineering Issues in Multi-Agent Transport Simulations

Keynote II Computational and Software Engineering Issues in Multi-Agent Transport Simulations

Modelling Inductive Charging of Battery Electric Vehicles using an Agent-Based Approach

The introduction of battery electric vehicles (BEVs) could help to reduce dependence on fossil fuels and emissions from transportation and as such increase energy security and foster sustainable use of energy resources. However a major barrier to the introduction of BEVs is their limited battery capacity and long charging durations. To address these issues of BEVs several solutions are proposed such as battery swapping and fast charging stations. However apart from these stationary modes of charging, recently a new mode of charging has been introduced which is called inductive charging. This allows charging of BEVs as they drive along roads without the need of plugs, using induction. But it is unclear, if and how such technology could be utilized best. In order to investigate the possible impact of the introduction of such inductive charging infrastructure, its potential and its optimal placement, a framework for simulating BEVs using a multi-agent transport simulation was used. This framework was extended by an inductive charging module and initial test runs were performed. In this paper we present the simulation results of these preliminary tests together with analysis which suggests that battery sizes of BEVs could be reduced even if inductive charging technology is implemented only at a small number of high traffic volume links. The paper also demonstrates that our model can effectively support policy and decision making for deploying inductive charging infrastructure.

BENCHMARKING MINIMUM PASSENGER WAITING TIME IN ONLINE TAXI DISPATCHING WITH EXACT OFFLINE OPTIMIZATION METHODS

This paper analyses the use of exact offline optimization methods for benchmarking online taxi dispatching strategies where the objective is to minimize the total passenger waiting time. First, a general framework for simulating dynamic transport services in MATSim (Multi-Agent Transport Simulation) is described. Next, the model of online taxi dispatching is defined, followed by a formulation of the offline problem as a mixed integer programming problem. Three benchmarks based on the offline problem are presented and compared to two simple heuristic strategies and a hypothetical simulation with teleportation of idle taxis. The benchmarks are evaluated and compared using the simulation scenario of taxi services in the city of Mielec. The obtained (approximate) lower and upper bounds for the minimum total passenger waiting time indicate directions for further research.

Heterogeneous Tolls and Values of Time in Multi-agent Transport Simulation

In evolutionary algorithms, agents’ genotypes are often generated by more or less random mutation, followed by selection based on the fitness of their phenotypes. This paper shows that elements of this principle can be applied in multi-agent transport simulations, in the sense that a router, when faced with complex interactions between heterogeneous toll levels and heterogeneous values of time, can resort to some amount of randomness rather than being able to compute the exact best solution in every situation. The computational illustrations are based on a real world case study in the province of Gauteng, South Africa.

An Alternative Approach to Network Demand Estimation: Implementation and Application in Multi-Agent Transport Simulation (MATSim)

This paper introduces a novel network demand estimation framework consistent with the input data structure requirements of Multi-Agent Transport Simulation (MATSim). The sources of data are the American Community Survey, US Census Bureau, National Household Travel Surveys, travel surveys from South East Florida Regional Planning Authority, OpenStreetMap and Florida Statewide Transportation Engineering Warehouse for Archived Regional Database. The developed framework employs mathematical and statistical methods to derive probability density functions and multinomial logit models for activity and location choices. The implementation of demand estimation process resulted into the creation of 1,200,889 agents (only those using cars). The scenario for the estimated agents was configured and simulated in MATSim. The results from the simulated scenario resulted in the expected morning, afternoon and evening traffic patterns as well as the desirable level of agreement between simulated and observed traffic volumes.

Including joint decision mechanisms in a multiagent transport simulation

In recent years, there has been a growing interest in the social dimension of travel, and how travel decisions are influenced not only by the global state of the transportation system, but also by joint decisions and interactions with social contacts. Empirical studies pointed out the relevance of certain types of social interactions to model travel behavior correctly: synchronization of household members has been identified as significantly influencing individual’s travel decisions, and friendships were shown to influence the choice of leisure locations.Multiagent approaches, by representing decision making at the level of the individual, seem to be one of the most promising approaches to take into account such interactions when forecasting.The work presented here extends a state of the art multiagent simulation software, MATSim, with capabilities to capture joint decisions. It consists of a general framework to allow arbitrary joint behaviors to be represented, without constraints on the topology of the social network. An implementation of this framework for the important case of intrahousehold ride sharing is proposed, and demonstrated on a simple test scenario.The results show that the extended process is able to find the expected state, and is mature for validation against travel diary data.

An agent-based cellular automaton cruising-for-parking simulation

This paper reports on the development of an agent-based cruising-for-parking simulation using the cellular automaton (CA) approach. The software was tested on a small-scale scenario, and a first verification step was performed for a real-world scenario for the town center of Zürich. Approaches to integrating the simulation into MATSim, a multi-agent transport simulation program, are discussed. The software is open source and can be downloaded from a free software repository. Empirical data that may be valuable for future model calibration are currently being surveyed in a global positioning system (GPS) study at the authors’ institute.