Multi-modal urban transit network design considering reliability: multi-objective bi-level optimization
Introduction
One of the major issues facing people and managers in urban life, especially in metropolitan areas, is traffic problems or, more generally, urban transportation problems. As the population grows and the number of vehicles increases, it becomes more and more difficult for citizens in large cities to travel. This leads to problems such as waste of effective time of citizens, fatigue, nervousness, noise and environmental pollution, loss of energy, and urban warming. Several solutions have been proposed to resolve these problems, such as culture building, management, deterrence (e.g., different traffic schedules and bans on vehicles on different days), etc. However, the prerequisite for all these solutions is the creation or improvement of transportation infrastructure so that it could provide the necessary and adequate capacity for the population of each city or region. Otherwise, from the transportation engineering perspective, if the appropriate basic infrastructure is not provided, none of the solutions will be able to solve the traffic problems. The preparation of the infrastructure required for the urban transportation system is commonly referred to as the urban transportation network design problem (UTNDP). UTNDPs cover a wide range but can be divided into two broad categories. In the first category, all or part of the urban transportation network is created. In the second category, some modifications are made to existing urban transportation networks, for instance, by increasing the capacity of links, converting two-way links into one-way links, etc. This paper focuses on changing the capacity of existing transportation networks to solve traffic problems by increasing the capacity of some links in urban transportation networks.
The main advantage of modeling in this research is to be able to design the urban transportation network of metropolitan areas, because urban transportation networks were mostly designed before the transformation into metropolitan areas and their network was traditionally expanded. Therefore, the redesign of the transportation network changes the residential, commercial, administrative, and other urban fabrics, and staggering costs are imposed on the transportation management. Hence, it is inevitable to make some decisions to solve the traffic dilemma, as the capacity of streets and roads does not meet the demand or traffic flow. In this way, not only the traffic problem will be solved under normal conditions, but also the congestion and heavy traffic will be prevented during the peak transport demand.
Transportation and traffic organizations in metropolitan areas are constantly trying to solve the problem of street and road traffic, because it has very serious consequences for the economy, culture, and society in a direct or indirect manner. One strategy is to build culture for transporting people by public transport. In any case, increasing the capacity of existing streets and roads is undeniable. Therefore, transportation management should increase the capacity of roads to meet the network demand and thus minimize the cost of this increase. In addition, the capacity of roads should not exceed a certain level.
To analyze the real-world problems, the functional results are obtained if the problem hypotheses can be defined close to the real world. One of the most important hypotheses is to consider different transport modes. In the real world, there are several modes whose demands are interrelated. The multi-modal network design problem (MMNDP) can well refer to the problems with at least two different modes for the UTNDP. The MMNDP can cover cars, taxis, trucks, buses, bicycles, motorcycles, subways, etc. The multi-modal problems in urban transportation networks often occur in the following forms [26]:
(a) Absence of interactions between flows of different modes; (b) Presence of interactions only between flows of different modes; (c) Presence of interactions between flows and decisions.
Another dimension of the MMNDP is the number of modes involved in the travel between two points. In this dimension, there are two cases: (a) The modes involved in the travel of passengers can be only one; (b) The modes involved in the travel of passengers can be more than one.
Certainty or uncertainty is among the effective hypotheses for bringing the problems closer to the real world. In the real world, there is no certainty, and this is also the case with urban transportation networks. In UTNDPs, there is uncertainty surrounding various problems, including travel time, demand, flow, and link capacity, which leads to uncertainty in the network and is called the reliability of urban transportation networks. There are three criteria for assessing the reliability of a transportation network: connectivity reliability, travel time reliability, and capacity reliability [15].
This was a brief description of urban transportation problems. These problems are essentially modeled at two levels, since there are two major decision-makers, namely the citizens and the urban transportation management. It is difficult to solve a bi-level network design problem by the exact solution method because the problem has NP-hardness. One of the initial studies on the bi-level problems was performed by Ben-Ayed and Blair [7] who concluded that even a simple bi-level problem with both linear up and down levels is still NP-hard. Another reason is the non-convexity of the bi-level network design problem. Even though both the up and down levels in the bi-level problems are convex, this does not guarantee the convexity of the bi-level problem [44].
Considering the nature of urban transportation problems, this paper considers two types of decision-makers for modeling. The first decision-maker is the management of the urban transportation network, which seeks to increase the capacity of the network links considering both public transport modes. This can improve the travel time reliability and network capacity reliability by increasing the capacity of the links. These modes of transport are completely independent in terms of flow through links, and passengers can use a combination of modes for their travel and switch transport modes at intersection nodes. The management also seeks to minimize the cost of increased capacity of the links. The result of reliability maximization is overall satisfaction, reduction in traffic, and reduction in travel time or time spent on travel. The second decision-maker in this problem is the passengers who choose a path with minimum average total travel time and maximum capacity reliability according to the management decisions. This problem considers uncertainty in total travel time, flow through links, and travel demand. The modeling also assumes that the decision-makers of the low-level problem are the passengers who want to travel during peak traffic in addition to the normal demand. Another important point to note in this paper is the correlation between all the links forming the path. In the real world, the links forming a path have a correlation due to the passing flow, which should be considered in the calculations. However, according to the hypotheses, there is no correlation between the links with different modes, because they have completely independent movement flows.
This paper consists of 7 sections. Section 2 presents the literature review on the subject of this paper. Section 3 describes a multi-level dual-mode model for the problem presented in this study. Then, Section 4 presents a meta-heuristic solution using the particle swarm optimization (PSO) algorithm for the model presented in Section 3. To apply the proposed model and algorithm, Section 5 examines part of urban transportation network in the metropolitan area of Tehran. Finally, Section 6 draws conclusions and makes suggestions for future research.
Section snippets
Literature review
According to the research area of this paper, which deals with the design of urban transportation networks in uncertain conditions, this section discusses recent studies conducted in the related fields.
In urban areas, travel times in multi-modal transportation networks vary due to random fluctuations in demand and deviations in supply. Many experimental studies have concluded that travel time uncertainty has a significant influence on passengers' behavior in the selection of mode and path [1,9,
Presentation of mathematical modeling
This study considers some hypotheses that should be mentioned for the readers before presenting the model.
- (a)
The modeling proposed in this study considers a dual-mode urban transportation network, where the flows passing through the modes are completely independent. However, passengers can change their transport modes during the travel. (In particular, two modes of public transport are considered in the presented case study: subway and bus.)
- (b)
The considered urban transportation network is available,
Solution approach
The main objective of this study is not to investigate the performance of the solution methods provided for the proposed model. The particle swarm optimization (PSO) algorithm not only has a low computational cost in terms of required memory and speed, but also has a very good homogeneity performance. In this regard, the proposed modeling is solved inspired by various studies on the application of PSO algorithm to bi-level problems, such as the studies by Han et al. [30] and Yang et al. [74].
Case study
In terms of population and population density, Tehran, the capital of Iran, is one of the Iran's largest metropolitan areas, consisting of 22 districts. In the central part of the city, the roads and streets were designed about a hundred years ago. Also, the concentration of public and private organizations in this part is very high. It is part of the traffic plan area, because the travel demand in this area is much higher than in other areas. This shows that it is impossible to create new
Conclusion
As shown in the previous section, the modeling presented in this paper can increase the travel time reliability and system capacity reliability to an acceptable level. From the results of this study, urban transportation management can improve public welfare and travelers' satisfaction and greatly reduce urban traffic and indirectly reduce energy consumption by increasing the capacity of transport lines. On the other hand, travelers can choose the paths that have less travel time, and the
CRediT authorship contribution statement
Amir Hossein Barahimi: Methodology, Investigation, Software, Writing – original draft. Alireza Eydi: Conceptualization, Supervision, Visualization, Writing – review & editing. Abdolah Aghaie: Methodology, Validation.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (80)
- et al.
Transit network design with allocation of green vehicles: a genetic algorithm approach
Transp Res Part C Emerg Technol
(2009) - et al.
Drivers’ willingness-to-pay to reduce travel time: evidence from the San Diego I-15 congestion pricing project
Transp Res Part A Policy Practice
(2003) - et al.
Stochastic multi-objective models for network design problem
Expert Syst Appl
(2010) - et al.
Route choice decision under travel time uncertainty
Transp Res Part A Policy Pract
(2005) - et al.
Transportation network reliability in emergency response
Transp Res Part E Log Transp Rev
(2015) - et al.
A review of urban transportation network design problems
Eur J Operat Res
(2013) - et al.
Estimating travel time distribution under different traffic conditions
Transp Res Proc
(2014) - et al.
A solution to bi/tri-level programming problems using particle swarm optimization
Inf Sci
(2016) - et al.
'The value of time and reliability: measurement from a value pricing experiment'
Transp Res Part E Logist Transp Rev
(2001) - et al.
Multi-scenario optimization approach for assessing the impacts of advanced traffic information under realistic stochastic capacity distributions
Transp Res Part C Emerg Technol
(2017)
Quantifying travel time variability at a single bottleneck based on stochastic capacity and demand distributions
J Intell Transp Syst
Network with degradable links: capacity analysis and design
Transp Res Part B Methodol
GIS based destination accessibility via public transit and walking in Auckland, New Zealand
J Transp Geogr
Shortest path problem considering on-time arrival probability
Transp Res Part B Methodol
Reliable route guidance: a case study from Chicago
Transp Res Part A Pol Pract
'Evaluating the reliability of a stochastic distribution network in terms of minimal cuts'
Transp Res Part E Log Transp Rev
Prediction of travel time variability for cost-benefit analysis
Transp Res Part A Policy Pract
Finding most reliable paths on networks with correlated and shifted log–normal travel times'
Transp Res Part B Methodol
First-best marginal cost toll for a traffic network with stochastic demand
Transp Res Part B
Modelling travel time reliability with the Burr distribution
Proc Soc Behav Sci
Reliability-based network flow estimation with day-to-day variation: a model validation on real large-scale urban networks
J Intell Transp Syst
High-resolution spatio-temporal modeling of public transit accessibility
Appl Geogr
Reserve capacity of a signal-controlled road network
Transp Res Part B Methodol
A reliability-based land use and transportation optimization model
Transp Res Part C Emerg Technol
A hybrid evolutionary algorithm for multiobjective variation tolerant logic mapping on nanoscale crossbar architectures
Appl Soft Comput
Modeling travel time distributions on a road network
'Bi-level multi-objective model for existing link capacity expansion problem across urban transportation network considering travel time reliability: presenting dynamic particle swarm algorithm'
Sādhanā
Calculation of urban transportation network reliability considering correlation among the links comprising a route
Scientia Iranica
Computational difficulties of bilevel linear programming
Oper Res
The benefits of a high-resolution analysis of transit accessibility
Int J Geograph Inf Sci
The multi-criteria road network design problem in an urban area
Transportation
A capacity related reliability for transportation networks
J Adv Transp
Effect of route choice models on estimating network capacity reliability
Trans Res Record J Transp Res Board
New reserve capacity model of signal-controlled road network
Transp Res Record J Transp Res Board
Transport network design problem under uncertainty: a review and new developments
Transp Rev
Reliable space–time prisms under travel time uncertainty'
Ann Assoc Am Geogr
Reliable shortest path problems in stochastic time-dependent networks
J Intell Transp Syst
A reliability-based network design problem
J Adv Transp
Cited by (21)
Relief network design problem (RNDP): A scoping review, challenges, and opportunities
2024, International Journal of Disaster Risk ReductionImproving access to urban parks through public transit optimization
2024, Frontiers of Architectural ResearchIntegrated scheduling method for fleet wave sorties and maintenance of naval distributed platforms
2024, Advanced Engineering InformaticsResilience evaluation of UAV swarm considering resource supplementation
2024, Reliability Engineering and System SafetyGlobalized robust bilevel optimization model for hazmat transport network design considering reliability
2023, Reliability Engineering and System SafetyAn approach for reliability optimization of a multi-state centralized network
2023, Reliability Engineering and System Safety