Simulation of Evacuation Route Choice

Available online at www.sciencedirect.com

ScienceDirect Transportation Research Procedia 20 (2017) 740 – 745

12th International Conference "Organization and Traffic Safety Management in Large Cities", SPbOTSIC-2016, 28-30 September 2016, St. Petersburg, Russia

Simulation of Evacuation Route Choice Vladimir Zyryanov1a*Anastasia Feofilova 1b* Don State Technical University, 1 Gagarin sq., Rostov-on-Don, 344000, Russia

Abstract In last years is significant attention to the simulation of transport systems for emergencies. Simulation provides different methods to decrease evacuation time when traffic demand can change very fast, some of sections of road network is destroyed and capacity is limited. For new traffic states, route guidance has to be considered on the base of route choice model. The following conceptual approaches are used as tools for these problems solution in our research: traffic monitoring; estimation of congestion conditions; traffic simulation on urban network and the forecast of the system efficiency of emergency evacuation. byby Elsevier B.V.B.V. This is an open access article under the CC BY-NC-ND license TheAuthors. Authors.Published Published Elsevier © 2017 2016The (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the 12th International Conference "Organization and Traffic Peer-review under responsibility of the organizing committee of the 12th International Conference “Organization and Traffic Safety Safety Management in Large Cities". Management in large cities” Keywords: Traffic simulation, traffic demand, emergency, route choice

1. Introduction Transportation evacuation is an important multi-disciplinary research topic that has received a great deal of attention from the research community and general public. Most of all researchers are aimed of preparing evacuation plans and applying traffic simulation models. Virginia P. Sisiopiku, Steven L. Jones et al. [Sisiopiku et al. (2004)] prepared report that shows how microscopic traffic simulation (with CORSIM) can be used to assist decision making for regional emergency preparedness through a series of case studies implemented on Birmingham’s regional transportation network. Gang-Len Chang et al. [Chang et al. (2007)] created traffic management system for the Eastern Shore region that enables responsible agencies to design the optimal routing plan and monitor traffic conditions under

* Corresponding author. Tel.: +0-000-000-0000 ; fax: +0-000-000-0000 . E-mail address: [email protected] a*, [email protected] b

2352-1465 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the 12th International Conference “Organization and Traffic Safety Management in large cities” doi:10.1016/j.trpro.2017.01.119

Vladimir Zyryanov and Anastasia Feofilova / Transportation Research Procedia 20 (2017) 740 – 745

different safe time windows during emergency evacuation. Y. Wang, C. Zhang [Wang and Zhang (2013)] presented a practical method to identify key nodes or links that are important to the network and to provide a roadmap for system deployment and integration. Also authors verifies that dynamic alternative route guidance techniques help maximize the evacuation efficiency of congestion routes. Ying Liu, Nan Zou, and Gang-Len Chang [Liu et al. (2005)] proposed an integrated evacuation system for Ocean City, Maryland to prepare for potential hurricane attacks. Their evacuation system designed for Ocean City, Maryland under hurricane attacks is embedded with six candidate plans. Each plan, specified for a predefined evacuation network, is obtained by adjusting optimized control strategies from the optimization module based on simulation results. Maria A. Konstantinidou et al. [Konstantinidou et al. (2014)] focused on efforts and models for the estimation and assessment of post-disaster network performance and on the problem of decision-making and planning of post-disaster network operations. Zockaie, Mahmassani, Saberi and Verbas [Zockaie and Mahmassani (2014)] explored some of the dynamics of urban network traffic flow during a large-scale evacuation in the context of the Network Fundamental Diagram. Frequent route switching by adaptive drivers can artificially increase the average network flow but does not necessarily increase the network output (trip completion rate). Adaptive driving increases fluctuations in the NFD; however, it reduces hysteresis and gridlock while increasing network capacity. Researchers found that the structure of the evacuation demand can significantly affect network performance, also that the linear relationship between average network flow and trip completion rate does not hold when a network is highly congested and under disruption, and sufficient number of adaptive drivers exist. Nirajan Shiwakoti et al. [Shiwakoti et al. (2013)] derived a summary of recent studies on emergency evacuation including the level of detail (micro, meso and macro), mode considered (auto, pedestrian, transit) along with features and findings. The authors found that it is apparent that all observed models, in one way or another, contain some deficit of analysis and knowledge. Commonly, this is in the form of neglecting whole modes of evacuation with specialized focus on specific modes (methods) and representation (scope) as well as lacking realistic assumption of travel behaviour. Analysis results of different investigations allowed us to make conceptual scheme of steps to preparing evacuation plan to prevent unnecessary traffic jams and save lives and properties during an emergency evacuation (see figure 1)

Fig.1. Conceptual Scheme to Preparing Evacuation Plan.

It is also interesting to note that the driving behaviour parameters like headway, acceleration, reaction time are adjusted for the evacuation and that the model structure and parameter settings are typically not changed [Tamminga et al. (2010)]. The following conceptual approaches are used as tools for these problems solution in our research: traffic monitoring; estimation of congestion conditions; traffic simulation on urban network and the forecast of the system efficiency of emergency evacuation. A part of our research is conducted on the basis of the AIMSUN model of Rostov-

741

742


on-Don CBD network. A set of simulations were carried out using various random seeds for the stochastic results. Database of traffic flow parameters is very important for accurate simulation. So we considered to use microscopic simulation to determine the effectiveness evacuation strategies taking into account the driver behaviour. 2. Simulation Framework To determine the effectiveness of the undertaken decision it would be appropriate to run several scenarios for the development of the traffic condition. Based on the drivers behaviour research in terms of their route choosing we propose the following classification of scenarios: 1. The driver will choose the alternative route 2. The driver probably will not choose an alternative route 3. The driver probably choose an alternate route To run these scenarios of drivers’ behavior we used the microsimulation software AIMSUN [Barcelo and Casas (2005), Sheffi (1985), Naumova and Zyryanov (2015)]. The objective of the simulation was to determine such parameters of route choice models at which the "behaviour" of vehicles will follow the above-stated scenarios (see Table 1). Table 1. Implementation of the Route Choice Models and their Parameters due to the Type of Scenarios. Simulation scenarios 1.The driver will choose the alternative route

Route choice models Proportional model Logit model С-Logit model

Recommended model parameters

D =[2;3] T =(60;100] T =[60;100] β does not influence

2. The driver probably will not choose an alternative route

Proportional model Logit model С-Logit model

3. The driver probably choose an alternate route

Proportional model Logit model С-Logit model

D =(0;0,5) T =(0;10) T =(0;10); β =(0,1; 1) D =[0,5;2) T =[10;60) T =(10;60)

β =0,5

The next objective of the simulation is the influence of the route choice models’ parameters on traffic conditions changing in the network. Firstly, we simulated a simple fragment of network with an opportunity to choice three different routes, then more complex network and finally a part of Rostov-on-Don network. In order to run various scenarios it is important to see how traffic and incidents correlate with route choice. Therefore, we analyze scenarios with incident on the one route. We found out that the drivers start using an alternate route only when the initial route LOS reaches F. In real conditions it corresponds to the situation when the driver, himself, sees in front of him congestion and if there is a physical possibility to change the route he uses an alternative path. Simulating a general drivers "obedience" (scenario 1), we found out that the initial route loading decreases more rapidly. In this case it may lead to LOS D on the alternative path and to LOS E in heavy traffic. Dealing with possible drivers routing (Scenarios 2 and 3) we did not observe the LOS increase in the initial route, detecting LOS E and LOS F there (see fig.2). We also found out that in case of heavy traffic the use of an alternative route does not lead to efficient results (there is still a congestion on the main route and traffic conditions on the alternate route R2 are getting worse) [Zyryanov and Feofilova (2016)].


743

Fig. 2. Traffic Conditions Changes in Scenarios Rank.

The next step of our researches was to study traffic conditions then the guidance acceptance parameter is changing. We defined with simulation scenarios there the guidance acceptance ranging from 0 to100 (step 10) that the following parameters’ groups can be distinguished for effective traffic management: x bad guidance acceptance (ga from 0 to 20) x enough guidance acceptance (ga from 20 to 70) x good guidance acceptance (ga from 70 to 100) Traffic speed changes on the original route taking into account the impact of the driver behavioral parameters presented on the figure 3. Figures above shows that to reduce the impact of traffic incidents it will be necessary to real-time traffic management. To implement these results was a part of transport planning and traffic simulation to prepare transport infrastructure of Rostov-on-Don for FIFA World Cup 2018. As an example of the dynamic traffic flow re-routing we studied the network area near the stadium functioning as the entrance to the city. For real data we used incident statistics in the period of 2010-2012 on the Eastern highway of Rostov-on-Don. The statistical data also provided the duration of incident consequences elimination.

Fig. 3. Speed – Time –Distance Graphs for Guidance Acceptance Calibration.

744


Strategy 0: Do nothing Strategy 1: 33 % (enough guidance acceptance) vehicles are re-routed from the route 1 to the route 2 Strategy 2: 33 % (enough guidance acceptance) vehicles are re-routed from the route 1 to the route 3 Strategy 3: drivers will not accept guidance rerouting. Force turning from route 1 to route 2, 3. For qualitative analysis of the strategies we used data obtained in AIMSUN software using application software (API) providing with the information on the vehicle speed at each of three routes at each designated time interval. Constructed space-time diagrams give a broad idea of the change in traffic conditions on the route while performing any impact.

Fig. 4. Speed – Time –Distance Graphs of all Modeled Strategies.

Quantitative analysis of the experimental results showed a decrease in the vehicles travel time using the strategy 1 on the initial route up to 200% and using strategy 2-3 up to 175%. 3. Conclusions The paper presents research aimed at finding and adapting models of routing traffic to the real conditions. By calibrating the model parameters we have determined that the forecast changes in the traffic situation should be under three scenarios: when the driver choose the recommended route, the driver will probably choose the route, the driver probably will not choose the route.


There were also defined following types of driver behavior: bad guidance acceptance, enough guidance acceptance, good guidance acceptance. Further research will continue for the following tasks: to predict the traffic situation in the development of transport infrastructure during process of evacuation planning; to study the traffic flows’ redistribution on the network in evacuation zones; to forecast the traffic situation in the case of incidents; to improve methods and algorithms for the dynamic traffic re-routing References Barcelo, J. and Casas J. (2005). Stochastic heuristic dynamic assignment based on AIMSUN microscopic traffic simulator. 85th Transportation Research Board 2006 Annual Meeting. Gang-Len Chang, Ying Liu, Haghani, A., Abbas M. A. (2007). An Integrated Real-Time Simulation/Optimization System for the Eastern Shore Highway Network: Research Methodology. Final Report, 124 p. Konstantinidou, M. A., Kepaptsoglou, K. L., Karlaftis, M.G. (2014). Transportation Network Post-Disaster Planning and Management: A Review, Part II: Decision-Making and Planning of Post-Disaster Operations. International Journal of Transportation, 2 (3):17–32. Naumova, N.A., Zyryanov, V.V. (2015). A method of computing the traffic flow distribution density in the network with new flow-forming objects being put into operation. Journal of Theoretical and Applied Information Technology, 78 (1): 76–83. Shiwakoti, N., Zhiyuan L., Hopkins, T., Young, W. (2013). An Overview on Multimodal Emergency Evacuation in an Urban Network. Australasian Transport Research Forum 2013 Proceedings, Australia. Sheffi, Y. (1985). Urban Transportation Networks: Equilibrium Analysis with Mathematical Programming Methods. Prentice-Hall, Englewood Cliffs, USA. Tu, H., Tamminga, G., Drolenga, H., De Wit, J., Van Der Berg, W. (2010) Evacuation plan of the city of Almere: simulating the impact of driving behaviors on evacuation clearance time. Procedia Engineering, (3): 67–75. Sisiopiku, V. P., Jones, S. L., Jr. Andrew J., Sullivan, S. S., Patharkar, X. T. (2004). Regional Traffic Simulation for Emergency Preparedness. Final Report, 64 p. Ying, L., Zou, N. and Gang-len Chang. (2005). An integrated emergency evacuation system for real-time operations. A case study of Ocean City, Maryland under hurricane attacks. In proceedings of 2005 IEEE Intelligent Transportation Systems, 464–469. Wang, Y., Zhang, C. (2013) Alternative route strategy for emergency traffic management based on its: a case study of xi’an ming city wall. Tehnički vjesnik 20(2): 359–364. Zockaie, A., Mahmassani, H., Saberi, M., and Verbas, O. (2014). Dynamics of Urban Network Traffic Flow during a Large-Scale Evacuation. Transportation Research Record: Journal of the Transportation Research Board, (2422): 21–33. Zyryanov, V., Feofilova, A. (2016). Evaluation Parameters of Re-routing Strategy Transport Research Arena (TRA) 2014 Proceedings, 10p. Available at : http://onlinelibrary.wiley.com/doi/10.1002/9781119307822.ch14/summary (viewed on: 12.04.2016).

745