DDNS Discrete Dynamics in Nature and Society 1607-887X 1026-0226 Hindawi Publishing Corporation 10.1155/2016/9848916 9848916 Research Article Boundedly Rational User Equilibrium with Restricted Unused Routes http://orcid.org/0000-0002-1543-5790 Sun Chao 1 Li Menghui 2 http://orcid.org/0000-0003-1617-0154 Cheng Lin 1 http://orcid.org/0000-0002-6138-5130 Zhu Senlai 1 http://orcid.org/0000-0002-6230-1955 Chu Zhaoming 3 Çinar Cengiz 1 School of Transportation Southeast University Nanjing 210096 China seu.edu.cn 2 School of Highway Chang’an University Xi’an 710064 China chd.edu.cn 3 Road Traffic Safety Research Center of the Ministry of Public Security Beijing 100062 China 2016 5122016 2016 11 07 2016 30 09 2016 06 11 2016 2016 Copyright © 2016 Chao Sun et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

A boundedly rational user equilibrium model with restricted unused routes (R-BRUE) considering the restrictions of both used route cost and unused route cost is proposed. The proposed model hypothesizes that for each OD pair no traveler can reduce his/her travel time by an indifference band by unilaterally changing route. Meanwhile, no route is unutilized if its travel time is lower than sum of indifference band and the shortest route cost. The largest and smallest used route sets are defined using mathematical expression. We also show that, with the increase of the indifference band, the largest and smallest used route sets will be augmented, and the critical values of indifference band to augment these two path sets are identified by solving the mathematical programs with equilibrium constraints. Based on the largest and smallest used route sets, the R-BRUE route set without paradoxical route is generated. The R-BRUE solution set can then be obtained by assigning all traffic demands to the corresponding generated route set. Various numerical examples are also provided to illustrate the essential ideas of the proposed model and structure of R-BRUE route flow solution set.

National Natural Science Foundation of China 51578150 51378119 Scientific Research Foundation of Graduate School of Southeast University YBJJ1679 Fundamental Research Funds for the Central Universities and the Research Innovation Program for College Graduates of Jiangsu Province KYLX15_0150 China Scholarship Council
1. Introduction

Perfect rationality is widely used in studying traditional transportation network models in which traveler always chooses the shortest (i.e., least utility) route, such as user equilibrium (UE ) and stochastic user equilibrium (SUE ) traffic assignment models. However, travelers may not always choose shortest route due to (1) lack of perfect travel information; (2) incapability of obtaining the shortest route with the complex traffic situations; and (3) certain “inertia” in decision making. Therefore people do not always choose the route with the maximum utility. They tend to seek a satisfactory route instead.

In the literature of evaluating habitual routes in route choice behavior, only 30% of respondents from Boston , 59% from Cambridge, Massachusetts , and 86.8% from Turin, Italy  chose the shortest routes. Based on GPS studies, Zhu  found that 90% of subjects in the Minneapolis-St. Paul region choose routes one-fifth longer than average commute time. All findings above revealed that people do not usually take the shortest routes and the used routes generally have higher costs than shortest ones.

It is more practical that traveler is boundedly rational (BR); traveler will not change his/her route if his/her travel time is a little longer than the shortest route. A series of experiments were conducted to empirically validate bounded rationality . The results showed that, in the repeated learning process, commuters would not change their routes unless the difference between preferred arrival time and actual arrival time exceeded a threshold. And boundedly rational route choice modeling observed from experiments provided a valid description of actual commuter daily behavior.

Simon, in 1957 , first proposed the notion of bounded rationality. And in 1987 , Mahmassani and Chang introduced it to traffic modeling. Since then, bounded rationality has received considerable attention in various transportation models, such as traffic safety , transportation planning [16, 17], traffic policy making , traffic assignment, and network design . All these studies indicated that traveler is boundedly rational in his/her decision-making process.

Boundedly rational user equilibrium (BRUE) is a network state such that travelers can take any route whose travel time is within a threshold of the shortest route time [22, 23, 2528]. Such a threshold is phrased by Mahmassani and Chang  as “indifference band.” In other words, no one can reduce his/her travel time by an indifference band by unilaterally changing his/her route. The indifference band is estimated from either laboratory experiment data or a behavioral study of road users (e.g., by surveys [7, 29]). By introducing indifference band for each OD pair, the BRUE relaxes UE assumption that travelers only take the shortest routes at equilibrium.

However, unlike the conventional UE model, the traffic flow under BRUE may not utilize any shortest or least-cost route; in another word the unused route cost may be lower than the used one. For example, the route flows are 0, 5, and 7 for three different routes on one OD pair, and the travel times are 10, 12, and 13, respectively. If the indifference band is 3, the above route flow solution is a BRUE solution. From the behavioral point of view, one might question the plausibility of this that the least travel time route has no traffic on it. Therefore, to make the model become behavioral more defensible, not only the used route cost should be restricted, but also the unused route cost should be restricted.

Uncertainties are unavoidable in transportation systems and make people become boundedly rational. Travelers do not know exactly the time that they arrive at the destination due to the travel time variability which is made by uncertainties. However, in many cases (such as going to work, having a meeting, and catching the train), travelers care more about arrive time than travel time; no one wants to be late; thus the largest and smallest route sets exist in travels’ trip process.

This paper makes contributions in three major areas: (1) considering both the restrictions of used route cost and unused route cost in the route decision process, we present a boundedly rational user equilibrium model with restricted unused routes (R-BRUE). This new model hypothesizes that for each OD pair no traveler can reduce his/her travel time by an indifference band by unilaterally changing route. Meanwhile, no route is unutilized if its travel time is lower than sum of indifference band and the shortest route cost. (2) We propose the largest and smallest used route sets which can be used to generate the used route set of R-BRUE model. These two used route sets are defined as the union and intersection of all R-BRUE solution set patterns, respectively. And (3) we develop two mathematical programs (MP) with equilibrium constraint to solve the critical values which is used to augment the largest and smallest used route sets.

The remainder of the paper is organized as follows. In Section 2, ε-R-BRUE (ε denotes the indifference band) is defined and its mathematical formulation is established. In Section 3, the largest and smallest ε-used route sets are defined, and their properties are studied. In Section 4, ε-R-BRUE route set without paradoxical route is generated. In Section 5, ε-R-BRUE route flow set without paradoxical route is constructed, and some examples are presented to illustrate the essential ideas of proposed model and the structure of R-BRUE route flow solution set. Finally, some conclusions and future work are provided.

2. Definition of <italic>ε</italic>-R-BRUE and Mathematical Formulation

In this section, we propose ε-boundedly rational user equilibrium model with restricted unused routes (R-BRUE) and the mathematical formulation of proposed model.

2.1. Definition of <bold> <italic>ε</italic></bold>-R-BRUE

Consider a transportation network G=N,A, where N and A denote the sets of nodes and links, respectively. Let W denote the set of OD pairs for which travel demand qω is generated between OD pair ωW, and let fkω denote the traffic flow on route kKω, where Kω is the set of routes connecting OD pair ω and all Kω constitute K. The feasible route flow set is to assign the traffic demand on the feasible routes: Ff:f0,kKωfkω=qω,ωW. Below is formal definition of ε-boundedly rational user equilibrium with restricted unused routes (R-BRUE).

Definition 1.

ε -boundedly rational user equilibrium with restricted unused routes (ε-R-BRUE) is a network state such that the travel cost of all used route is less than or equal to the sum of given indifference band ε=εωωW0 and the shortest route cost; meanwhile, the travel cost of the unused route is greater than or equal to the sum of ε and the shortest route cost; that is,(1)fkω>0CkωfminiKωCiωf+εω,kKω,ωW,fkω=0CkωfminiKωCiωf+εω,kKω,ωW,where f is the vector form of traffic flow fkω:f=,fkω,T and Ckω· is the route cost function on route k between OD pair ω.

We should point out that boundedly rational user equilibrium (BRUE) model only considers fkω>0CkωfminiKωCiωf+εω which do not take the cost of unused route into consideration. We first use fkω=0CkωfminiKωCiωf+εω to restrict the “irrational solutions.”

Equation (1) gives a necessary condition judging whether a flow pattern is R-BRUE and is equivalent to the following condition:(2)Ckωf>miniKωCiωf+εωfkω=0,Ckωf<miniKωCiωf+εωfkω>0.

In other words, a used route has lower cost than an unused one, which is the same as that in the UE (user equilibrium) setting. When εω=0 for each ω, the R-BRUE definition is reduced to the UE problem.

Theorem 2.

Any ε-R-BRUE solution is also a ε-BRUE solution. ε-BRUE solution may not, however, necessarily fulfill ε-R-BRUE conditions.

Proof.

let f be a route flow pattern to ε-R-BRUE model. Then, for fkω>0, CkωfminiKωCiωf+εω hold for all kKω and ωW; that is, fkω>0CkωfminiKωCiωf+εω, kKω, ωW, which satisfies ε-BRUE model.

For the converse situation, suppose that a flow allocation satisfies ε-BRUE conditions and in addition has an unused route which has a cost less than the sum of ε and the shortest route. Then ε-R-BRUE conditions are violated.

Usually ε-R-BRUE is nonunique. Denote a set containing all route flow patterns satisfying Definition 1 as ε-R-BRUE route flow solution set:(3)FR-BRUEεfF:fkω>0CkωfminiKωCiωf+εω,fkω=0CkωfminiKωCiωf+εω,kKω,ωW.

Theorem 3.

If the link cost function is continuous, ε-R-BRUE solution (ε0) is nonempty.

Proof.

First, Patriksson  showed that, when the link cost function is continuous, UE solution exists. Let fUEFUE be one UE route flow pattern, and set f1fi:fifUE,fi>0, f0fi+δ:fifUE,fi=0,Ckωfi<miniKωCiωfUE+εω, where δ is a very small positive parameter. Let f=f0,f1, when ε0,(4)fkω>0Ckωf<miniKωCiωf+εωminiKωCiωf+εω,fkω=0CkωfminiKωCiωf+εω,kKω,ωW.So f is ε-R-BRUE solution (ε0); that is, fFR-BRUEε. Given the continuous link cost function, at least one ε-R-BRUE flow pattern exists, and therefore FR-BRUEε.

2.2. R-BRUE Mathematical Formulation

We use slack variables ρkω to define R-BRUE mathematically. f is a R-BRUE distribution if and only if there exists πω whose physical meaning is the minimum route cost for every ω such that(5)Ckωf-πω-ρkω=0,kKω,ωW,fkωεω-ρkω=0,ifεωρkω,>0,o.w.,kKω,ωW,kKωfkω=qω,ωW,fkω0,kKω,ωW,ρkω0,kKω,ωW,where qω is the traffic demand between OD pair ω. Note that when ε=0 for all ω, (5) reduces to 0fkωCkωf-πω0, kKωfkω=qω which is the conventional UE conditions.

3. Largest and Smallest <italic>ε</italic>-Used Route Sets

In this section, we give the definition of the largest and smallest ε-used route sets, and we also discuss the properties of two proposed route sets.

3.1. Definitions

Here we give three definitions of used route set, largest ε-used route set (ε-URS) and smallest ε-URS of ε-R-BRUE model as follows.

Definition 4.

Given ε-R-BRUE route pattern fFR-BRUEε, the used route carries flow, while its travel cost is within the shortest cost plus the indifference band; that is,(6)aεf=kKω:fk>0,CkωfminiKωCiωf+εω,ωW.

Definition 5.

A largest ε-URS contains all used routes for every flow pattern in ε-R-BRUE flow set, mathematically:(7)Klε=fFR-BRUEεaεf.

Definition 6.

A smallest ε-URS contains the used routes which all flow patterns in ε-R-BRUE flow set have, mathematically:(8)Ksε=fFR-BRUEεaεf.

3.2. Monotonically Nondecreasing Largest <bold> <italic>ε</italic></bold>-URS

In the following, we will discuss the impact of the value of ε on the size of the largest ε-URS.

Theorem 7.

If 0ε<ε, then KlεKlε, where Klε is defined in (7).

Proof.

Let fεFM-BRUEε be one ε-R-BRUE route flow pattern, and set f1fi:fifε,fi>0, f0fi+δ:fifε,fi=0,Ckωfi<miniKωCiωfε+ε, where δ is a very small positive parameter. Let fε=f0,f1; then(9)fkω>0Ckωf<miniKωCiωf+εωminiKωCiωf+εω,fkω=0CkωfminiKωCiωf+εω,kKω,ωW.So fε is ε-R-BRUE solution, and aεfaεf. Based on (7), we can get that KlεKlε.

3.3. Paradoxical Route and Property of Smallest <bold> <italic>ε</italic></bold>-URS

We will give the definition of paradoxical route first.

Definition 8.

With the increasing of OD demand, the traffic flow on route p first increases and then decreases at the UE state. And the route p is defined as a paradoxical route.

This phenomenon contradicts our intuition. Following that, we use an example to illustrate this paradox.

The network topology of the test network, link, and traffic demand characteristics are depicted in Figure 1. The network consists of 4 nodes, 5 links, and 1 OD pair. Red curves on the right indicate 3 routes.

Test network and routes characteristics.

The equilibrium route flow f1 by varying the traffic demand q from zero to infinity is shown in Figure 2. When the traffic demand is lower than 4, all the travelers pick route 1; and f1 is increasing with the increasing of demand. While q reached to 4, f1 is decreasing with the increasing of q. And when the demand reaches to 24, the flow on route 1 becomes 0. Based on Definition 8, we can get that route 1 is the paradoxical route.

The UE traffic flow of route 1 under different demand.

When 0ε<ε, KsεKsε, where the network contains paradoxical route, monotonically nondecreasing property of smallest ε-URS may not be satisfied due to exiting paradoxical route in the network. Consider the test network in Figure 3 which consists of 4 nodes, 6 links, and 1 OD pair. The link travel time functions and traffic demand are reported in Figure 3. Red curves on the right indicate 4 routes.

Test network and routes characteristics.

Solving the UE where ε=0 for the R-BRUE, we have f1=0.25, f2=11.25, f3=11.25, and f4=9.25. Therefore, the smallest ε-URS is Ks0=1,2,3,4. When ε=48, f1=0, f2=15, f3=15, and f4=2, the travel time is C1=60, C2=57, C3=57, and C4=10. The minimum OD travel time is 10 and all the utilized routes have travel times of no more than 10 + 48 = 58. Therefore, the route flow pattern is a valid R-BRUE flow. However, the smallest ε-URS which is Ks48=2,3,4 does not satisfy the monotonically nondecreasing property.

Theorem 9 (monotonically nondecreasing smallest <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M147"><mml:mrow><mml:mi mathvariant="bold">ε</mml:mi></mml:mrow></mml:math></inline-formula>-URS without paradoxical route).

If 0ε<ε, then KsεKsε for a network without paradoxical route, where Ksε is defined in (8).

Proof.

Assume that there exist two bands ε and ε which satisfy that 0ε<ε such that KsεKsε. Then, there must exist a route k such that kKsε and kKsε. In another word, fkε0 for all ε-R-BRUE route flow patterns, and fkε can be equal to 0 for some ε-R-BRUE route flow patterns. And this contradicts the assumption that the network has no paradoxical route.

When ε varies from zero to infinity, the minimum number of routes the largest ε-URS and smallest ε-URS without paradoxical route contains is the UE shortest routes when ε=0; that is, Kl0=Ks0KUE. The maximum number of routes the largest ε-URS and smallest ε-URS without paradoxical route contains is all feasible routes, meaning all feasible routes will be utilized if the indifference band is too large. Then we have KUEKlεK and KUEKsεK.

Given ε, the largest ε-URS (defined in (7)) is a set of all used routes under ε-R-BRUE set. And the smallest ε-URS (defined in (8)) is a set of the used routes which must have traffic flows under all ε-R-BRUE set. It is possible that some used routes for one ε-R-BRUE flow pattern are not used for other flow patterns and vice versa. This necessitates the exploration of the interior structure of ε-R-BRUE route set. Theorems 7 and 9 provide us with one approach of analyzing the structure of ε-R-BRUE without paradoxical route by varying values of ε.

4. Generation of <italic>ε</italic>-R-BRUE Route Set without Paradoxical Route

The route set is finite, while ε is treated as a continuous parameter for the time being. Starting with the UE route set when ε=0, provided the network topology and the link cost functions, UE can be determined by some established algorithms, for example, simplicial decomposition with disaggregated (DSD) algorithm , gradient projection algorithm , or origin-based algorithm . According to Theorems 7 and 9, when ε is gradually increased, more routes will be included in the largest and smallest ε-URS, and we should be able to identify those used routes one by one, until all alternative routes are included. This offers the theoretical foundation for deriving different combinations of used routes by varying ε subsequently.

4.1. Definition of Critical Points in <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M185"><mml:mrow><mml:mi>ε</mml:mi></mml:mrow></mml:math></inline-formula>-R-BRUE

It is assumed that there are n alternative routes for OD pair ω; that is, K=1,,n and K=n, where K is the cardinality of set K. Among these n routes, there are r shortest routes at the UE; that is, KUE=1,,r and KUE=rn. Below are the definitions of critical points of the largest and smallest ε-URS in R-BRUE without paradoxical route.

Definition 10.

The largest/smallest ε-URS will remain the same until ε reaches a special value, and we define this value as critical points of the largest/smallest ε-URS for OD pair Omega; that is,(10)εl,1infε>0KUEKlε;εl,jinfε>0Klεl,j-1Klε,εl,Jinfε>0Klε=K.(11)εs,1infε>0KUEKsε;εs,iinfε>0Ksεs,i-1Ksε;εs,Iinfε>0Ksε=K,where j=1,,J, i=1,,I are the unique sequences of finite critical points εl and εs, with εl,0=εs,0=0, εl,J+1=εs,I+1=.

We give the physical meaning of εl and εs. A “newly added route” of the largest ε-URS is defined as the route which is unavailable under εl,j-1 but available when ε=εl,j. And a “newly added route” of the smallest ε-URS is defined as the route which is available under εs,z:zi for all route flow patterns, but unavailable under εs,i-1 for some route flow patterns. We can define their “newly added route” as rjkK:kKlεl,j,kKlεl,j-1 and tikK:kKsεs,i,kKsεs,i-1, respectively. We should notice that the number of “newly added routes” may be two or more at the same time.

We use Figure 4 to more intuitively illustrate the definition of critical points of the largest and smallest ε-URS in R-BRUE without paradoxical route.

Monotonically nondecreasing property of critical points.

Clearly, from Figure 4, we can observe that the largest and smallest ε-URS can be described as Klε=KUE,Klε1,,KlεP and Ksε=KUE,Ksε1,,KsεQ with a fixed indifference band ε, where εl,Mε<εl,M+1, εs,Nε<εs,N+1, P=minM,J, and =minN,I.

4.2. <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M226"><mml:mrow><mml:mi>ε</mml:mi></mml:mrow></mml:math></inline-formula>-R-BRUE Route Set without Paradoxical Route for One OD Pair

Definition 10 says that the largest and smallest ε-URS in R-BRUE without paradoxical route includes more routes when ε increases to some critical values. Thus, the mathematical programs (MP) with equilibrium constraint can be developed to solve these critical values in largest and smallest ε-URS. Below is the MP equation for calculating largest ε-URS:(12)minεl,j(13a)s.t.Ckf-π-ρk=0,kK,(13b)fkεl,j-ρk=0,ifεl,jρk,>0,o.w.,kK,(13c)kKfk=q,(13d)fk0,kK,(13e)ρk0,kK,(13f)jKlεl,j-1fj<q,where j=1,,J. Equations (13a)–(13e) are to guarantee the route flow pattern is a feasible R-BRUE; (13f) tries to push a small amount of flow from the largest used route set Klεl,j-1 to some newly largest used route if εl is increased a little bit. When (13a), (13b), (13c), (13d), (13e), and (13f) are solved, optimal solutions f,εl,j,π will be obtained. And the newly added route rj can be derived from the traffic flow f.

For any route iK, following MP equation is used to get the smallest ε-URS:(14)maxεs,i(15a)s.t.Ckf-π-ρk=0,kK,(15b)fkεs,i-ρk=0,ifεs,iρk,>0,o.w.,kK,(15c)kKfk=q,(15d)fk0,kK,(15e)ρk0,kK,(15f)fi=0.Equations (15a)–(15e) are to guarantee the route flow pattern is a feasible R-BRUE; and (15f) is to insure that no traffic flow travels on route i. When εs,i for all routes are worked out, εs,i are sorted by their size. Then we get the form of εs,i as (11). When (15a), (15b), (15c), (15d), (15e), and (15f) are solved, optimal solutions f,εs,i,π will be obtained. And the newly added route ti can be derived from the traffic flow f.

Equations (13b), (13f), and (15b) are inequalities without equal sign. A small positive parameter δ (such as 0.01) is introduced to deal with this problem. And (13b), (13f), and (15b) are replaced by (16)fkεl,j-ρk=0,ifεl,jρk,δ,o.w.,-jKlεl,j-1fjδ-1q,fkεs,i-ρk=0,ifεs,iρk,δ,o.w.Equations (13a), (13b), (13c), (13d), (13e), (13f), (15a), (15b), (15c), (15d), (15e), and (15f) can be solved by GAMS software [25, 34].

4.3. Instance of <bold> <italic>ε</italic></bold>-R-BRUE Route Set without Paradoxical Route

The network topology of the test network, travel time functions, and traffic demand are depicted in Figure 5. The network consists of 3 nodes, 5 links, and 1 OD pair. Red curves on the right indicate 4 routes. The UE is f1=10, f2=10, f3=0, and f4=0.

Test network and routes characteristics.

Solving (13a), (13b), (13c), (13d), (13e), (13f), (15a), (15b), (15c), (15d), (15e), and (15f), we have the largest critical values εl,0=0, εl,1=1, εl,2=1.1, and εl,3= and the smallest critical values εs,0=0, εs,1=2.85, εs,2=4.9, and εs,3=. There are three cases for the largest ε-URS:

0ε<1:Klε=1,2, and when ε=0, f=10,10,0,0, C=10,10,10,11.1.

1ε<1.1:Klε=1,2,3, and when ε=1, f=10,10,0+,0, C=10,10,10,11.1.

ε1.1:Klε=1,2,3,4, and when ε=1.1, f=10,10,0+,0+, C=10,10,10,11.1.

And there are three cases for the smallest ε-URS:

0ε<2.85:Ksε=1,2, and when ε=0, f=10,10,0,0, C=10,10,10,11.1.

2.85ε<4.9:Ksε=1,2,4, and when ε=2.85, f=11.1,7.6,1.3,0+, C=11.1,8.25,11.1,11.1.

ε4.9:Ksε=1,2,3,4, and when ε=4.9, f=7.1,12,0+,0.9, C=7.1,12,12,12.

All largest and smallest ε-URS are also illustrated in Figure 6. For route 3, we can see that when the critical value ε reaches 1, route 3 will join in the used route set for some route flow patterns; and when ε reaches 4.9, route 3 will be the used route for all route flow patterns. In other words, the yellow bar in Figure 6 means the route is either the used route or unused route in different route flow patterns, while the blue bar represents the route must be a used route for all route flow patterns. If ε is calibrated from empirical data as 3, then Kl3=1,2,3,4, Ks3=1,2,4. Therefore, routes 1, 2, and 4 must be the used routes, and route 3 may carry traffic flow or not. The used route can be described as K3=1,2,4,1,2,3,4.

ε -URS under all the critical points.

4.4. <bold><italic>ε</italic></bold>-R-BRUE Route Set without Paradoxical Route for Multiple OD Pairs

For a network with total W OD pairs, let εl,jω and εs,iω be the largest and smallest critical points for OD pair ωW, j=0,1,,Jω, i=0,1,,Iω, respectively. Then, εl,jεl,jω and εs,iεs,iω are the sets of largest and smallest critical points for all OD pairs.

For one OD pair vW, we also can use (13a), (13b), (13c), (13d), (13e), (13f), (15a), (15b), (15c), (15d), (15e), and (15f) to solve the largest and smallest critical points. The only difference is route costs need the information of route flows across all OD pairs. And route flows fω, ωW, ωv are parameters when calculating the critical points of v. Then, we modify (13a), (13b), (13c), (13d), (13e), (13f), (15a), (15b), (15c), (15d), (15e), and (15f) as follows:(17)minεl,jvs.t.Ckvf-πv-ρkv=0,kKv,fkvεl,jv-ρkv=0,ifεl,jvρkv,>0,o.w.,kKω,kKωfkω=qω,ωW,fkv0,kKv,ρkv0,kKv,jKlεl,j-1vfjv<qv,maxεs,ivs.t.Ckvf-πv-ρkv=0,kKv,fkvεs,iv-ρkv=0,ifεs,ivρkv,>0,o.w.,kKω,kKωfkω=qω,ωW,fkv0,kKv,ρkv0,kKv,fiv=0.

Then the same approach can be used to generate the ε-R-BRUE route set without paradoxical route for multiple OD pairs.

By far we have proposed how to solve ε-R-BRUE route set without paradoxical route for both single OD pair and multiple OD pairs. The following will discuss the methodology of constructing ε-R-BRUE route flow set without paradoxical route.

5. Construction of <italic>ε</italic>-R-BRUE Route Flow Set without Paradoxical Route

We have already analyzed the interior structure of ε-R-BRUE route set without paradoxical route last section. As the indifference band gradually increases, more routes will begin to carry flows (the largest R-BRUE route set), and more routes must carry flows (the smallest R-BRUE route set). Based on this characteristic, we decompose ε-R-BRUE route set into small subsets which are easier to study.

5.1. <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M315"><mml:mrow><mml:mi>ε</mml:mi></mml:mrow></mml:math></inline-formula>-R-BRUE Flow Set without Paradoxical Route for One OD Pair

Define a sequence of sets Fkε, k=0,,M-N, where M and N are the cardinalities of largest and smallest ε-URS. Then we assign all the traffic demands as follows:(18)F0εfF:i1,i2Ksε:fi1,fi2>0,Ci1-Ci2ε,fi=q;jKlεi:fj=0,Cj-minCiε,F1εfF:hKlεKsε;p1,p2Ksε,h:fp1,fp2>0,Cp1-Cp2ε,fp=q;jKlεp:fj=0,Cj-minCpε,h=Kl,1ε,,Kl,M-NεKsε,FkεfF:h1,,hkKlεKsε;p1,p2Ksε,h1,,hk:fp1,fp2>0,Cp1-Cp2ε,fp=q;jKlεp:fj=0,Cj-minCpε,h1,,hk:everykroutesinKl,1ε,,Kl,M-NεexcludeKsε,k=0,,M-N.

We use a bridge network to illustrate (18) in detail. The network topology of the test network, travel time functions, and traffic demand are depicted in Figure 7. The network consists of 4 nodes, 6 links, and 1 OD pair. Red curves on the right indicate 4 routes. The indifference band ε is set as 4, 10, 15, 30, and 60.

A bridge network and routes characteristics.

The UE is f1=10, f2=0, f3=0, and f4=10. That is, routes 1 and 4 are utilized under UE. Substitute Klεl,0=Ksεs,0=1,4, route costs, and the demand into (13a), (13b), (13c), (13d), (13e), and (13f), we obtain the largest critical values εl,0=0, εl,1=5, εl,2=11.3, and εl,3=. Solving (15a), (15b), (15c), (15d), (15e), and (15f), then we obtain the smallest critical values εs,0=0, εs,1=17.5, εs,2=52.5, and εs,3=. There are three cases for the largest ε-URS:

0ε<5:Klε=1,4, and when ε=0, f=10,0,0,10, C=50,65,55,50.

5ε<11.3:Klε=1,3,4, and when ε=5, f=10,0,0+,10, C=50,65,55,50.

ε11.3:Klε=1,2,3,4, and when ε=11.3, f=9.26,0+,2.59,8.15, C=49.63,60.93,60.93,49.63.

And there are three cases for the smallest ε-URS:

0ε<17.5:Ksε=1,4, and when ε=0, f=10,0,0,10, C=50,65,55,50.

17.5ε<52.5:Ksε=1,3,4, and when ε=17.5, f=7.5,0,0+,12.5, C=40,60,57.5,57.5.

ε52.5:Ksε=1,2,3,4, and when ε=52.5, f=17.5,0+,0+,2.5, C=80,80,47.5,27.5.

Then we assign all the traffic demands with diffident ε as follows.

When ε=4, Kl4=Ks4=1,4, and M=N=2, then (19)F04fF:f1,f4>0,C1f-C4f4,f1+f4=20.

When ε=10, Kl10=1,3,4, Ks10=1,4, M=3, and N=2, then (20)F010fF:f1,f4>0,C1f-C4f10,f1+f4=20;f3=0,C3f-minC1f,C4f10,F110fF:i1,i21,4,h3;p1,p21,3,4:fp1,fp2>0,Cp1f-Cp2f10,f1+f3+f4=20.

When ε=15, Kl15=1,2,3,4, Ks15=1,4, M=4, and N=2, then(21)F015fF:f1,f4>0,C1f-C4f15,f1+f4=20;j2,3:fj=0,Cj-minC1f,C4f15,F115fF:h2,3;p1,p21,4,h:fp1,fp2>0,Cp1f-Cp2f15,fp=20;j1,2,3,4p:fj=0,Cj-minCp15,h=2,3,F215fF:h1,h22,3;p1,p21,2,3,4:fp1,fp2>0,Cp1f-Cp2f15,fp=20.

When ε=30, Kl30=1,2,3,4, Ks30=1,3,4, M=4, and N=3, then(22)F030fF:i1,i21,3,4:fi1,fi2>0,Ci1f-Ci2f30,f1+f3+f4=20;f2=0,C2f-minCi30,F130fF:h2;p1,p21,2,3,4:fp1,fp2>0,Cp1f-Cp2f30,f1+f2+f3+f4=20.

When ε=60, Kl60=Ks60=1,2,3,4, and M=N=4, then (23)F060fF:i1,i21,2,3,4:fi1,fi2>0,Ci1f-Ci2f60,f1+f2+f3+f4=20.

For convenience, we only show the R-BRUE solution sets with ε=4,10,15 in Figure 8. Due to the flow conservation of the fixed demand, its R-BRUE solution sets can be characterized by routes 1, 3, and 4 in Figures 8(b) and 8(c). Figure 8(a) is composed of a 2-route green subset; Figure 8(b) is composed of a 2-route green subset and 3-route yellow subset; Figure 8(c) is composed of a 2-route green subset, 3-route yellow subset, and 4-route blue subset. The magenta legend denotes that the solutions do not satisfy ε-R-BRUE.

R-BRUE solution sets illustration.

R-BRUE solution set with ε=4

R-BRUE solution set with ε=10

R-BRUE solution set with ε=15

From Figure 8, we can see that (1) when ε=15, in F015 (green subset), only routes 1 and 4 carry flow, so the subset is a line. In F115 (yellow subset), route 3 begins to carry flow and f3>0. In F215 (blue subset), route 2 begins to carry flow and f2>0; (2) the solution set is bounded; this is because FR-BRUEεF:Ff:f0,kKωfkω=qω,ωW, while its closeness cannot be guaranteed due to Ckωf<miniKωCiωf+εωfkω>0. In Figure 8(b), the travel time of the UE solution fUE=10,0,0,10 on the magenta line is CUE=50,65,55,50, and C3=55<50+15, so f3 should be greater than 0 in R-BRUE model; hence CUE do not satisfy 10-R-BRUE; (3) without considering the boundary point of ε-R-BRUE, the solution set is monotonically nondecreasing with the increase of ε; and (4) the subset is not necessarily convex even though the link performance function is affine linear; this is because the constraint condition of Cj-minCiε is nonconvex. This property can be seen in the yellow block which is nonconvex in Figure 8(c). Hence, we cannot guarantee the convexity of ε-R-BRUE solution set.

5.2. <bold><italic>ε</italic></bold>-R-BRUE Flow Set without Paradoxical Route for Multiple OD Pairs

In the last section, we obtain ε-R-BRUE route set without paradoxical route for multiple OD pairs. Then, it is not difficult to generalize the methodology of constructing ε-R-BRUE flow set without paradoxical route for a single OD pair to multiple OD pairs. We assign all the traffic demands to the routes as follows:(24)F0εfF:ωW,i1ω,i2ωKsεω:fi1ω,fi2ω>0,Ci1ω-Ci2ωεω,fiω=qω;jωKlεωiω:fjω=0,Cjω-minCiωεω,FkεfF:ωW,h1ω,,hkωωKlεωKsεω;p1ω,p2ωKsεω,h1ω,,hkωω:fp1ω,fp2ω>0,Cp1ω-Cp2ωεω,fiω=qω;jωKlεωpω:fjω=0,Cjω-minCpωεω,h1ω,,hkωω:everykωroutesinKl,1εω,,Kl,M-NεωexcludeKsεω,kω=k,k=0,,M-N.

6. Conclusions

In this paper, a boundedly rational user equilibrium model with restricted unused routes (R-BRUE) was proposed. This new model assumes that travel cost of all used route is less than or equal to the sum of given indifference band and the shortest route cost; meanwhile, the travel cost of the unused route is greater than or equal to the sum of indifference band and the shortest route cost. The mathematical formulation of the proposed model was then established.

Before constructing the R-BRUE flow set, largest and smallest used route sets were explored first. As the value of the indifference band increases, some routes which were not utilized before will be taken, and thus the route set that contains the equilibrium flow was named as the largest used route set. As the value of the indifference band increases, some routes which must be utilized for all R-BRUE flow patterns will be taken, and this route set was defined as the smallest used route set. Paradoxical route is defined as that with the increasing of OD demand; the UE flow on the paradoxical route increases first and then decreases. The monotonically nondecreasing property of largest and smallest used route sets without paradoxical route is proved.

The critical values of the indifference band to augment the largest and smallest used route sets can be identified by sequentially solving a class of mathematical programs. After the largest and smallest used route sets are obtained, the whole R-BRUE flow set can be obtained by assigning all traffic demands to the corresponding generated route set.

The proposed model is appealing in modeling realistic travel behavior. But due to the nonconvexity of the feasible region, it is difficult to get the solution using mathematical programming method, and our method which is separation of the solution sets takes heavy computational burdens. In future research, we will study the solution algorithm and then apply the built model to real traffic network, such as the autonomous vehicles. Also, it is worthwhile to extend the proposed model to the network design problem, such as enhancing capacities of the established links, congestion pricing, and adding new links to an existing road network.

Competing Interests

The authors declare that they have no competing interests.

Acknowledgments

This research is supported by the National Natural Science Foundation of China (no. 51578150 and no. 51378119), the Scientific Research Foundation of Graduate School of Southeast University (no. YBJJ1679), the Fundamental Research Funds for the Central Universities and the Research Innovation Program for College Graduates of Jiangsu Province (no. KYLX15_0150), and the China Scholarship Council (CSC) Program sponsored by the Ministry of Education in China.

Wardrop J. G. Road paper. Some theoretical aspects of road traffic research ICE Proceedings: Engineering Divisions. Thomas Telford 1952 1 3 325 362 Daganzo C. F. Sheffi Y. On stochastic models of traffic assignment Transportation Science 1977 11 3 253 274 10.1287/trsc.11.3.253 2-s2.0-0017525664 Ramming M. S. Network Knowledge and Route Choice 2001 Massachusetts Institute of Technology Bekhor S. Ben-Akiva M. E. Ramming M. S. Evaluation of choice set generation algorithms for route choice models Annals of Operations Research 2006 144 235 247 10.1007/s10479-006-0009-8 MR2257910 2-s2.0-33747813444 Prato C. G. Bekhor S. Applying branch-and-bound technique to route choice set generation Transportation Research Record 2006 1985 19 28 2-s2.0-33947609432 Zhu S. The Roads Taken: Theory and Evidence on Route Choice in the Wake of The I-35W Mississippi River Bridge Collapse and Reconstruction 2010 University of Minnesota Mahmassani H. S. Chang G.-L. On boundedly rational user equilibrium in transportation systems Transportation Science 1987 21 2 89 99 10.1287/trsc.21.2.89 2-s2.0-0023349684 Mahmassani H. S. Jayakrishnan R. System performance and user response under real-time information in a congested traffic corridor Transportation Research Part A: General 1991 25 5 293 307 10.1016/0191-2607(91)90145-G 2-s2.0-44949274427 Jayakrishnan R. Mahmassani H. S. Hu T.-Y. An evaluation tool for advanced traffic information and management systems in urban networks Transportation Research Part C 1994 2 3 129 147 10.1016/0968-090X(94)90005-1 2-s2.0-0000006909 Hu T.-Y. Mahmassani H. S. Day-to-day evolution of network flows under real-time information and reactive signal control Transportation Research Part C: Emerging Technologies 1997 5 1 51 69 10.1016/s0968-090x(96)00026-5 2-s2.0-0031070616 Mahmassani H. S. Liu Y.-H. Dynamics of commuting decision behaviour under advanced traveller information systems Transportation Research Part C: Emerging Technologies 1999 7 2-3 91 107 10.1016/s0968-090x(99)00014-5 2-s2.0-0033117197 Srinivasan K. K. Mahmassani H. S. Role of congestion and information in trip-makers' dynamic decision processes: experimental investigation Transportation Research Record 1999 1676 44 52 2-s2.0-0033343046 Chen R. Mahmassani H. S. Travel time perception and learning mechanisms in traffic networks Transportation Research Record 2004 1894 209 221 2-s2.0-19944378804 Simon H. A. Models of Man: Social and Rational; Mathematical Essays on Rational Human Behavior in Society Setting 1957 New York, NY, USA John Wiley & Sons Sivak M. How common sense fails us on the road: contribution of bounded rationality to the annual worldwide toll of one million traffic fatalities Transportation Research Part F: Traffic Psychology and Behaviour 2002 5 4 259 269 10.1016/s1369-8478(03)00003-2 2-s2.0-0038458915 Khisty C. J. Arslan T. Possibilities of steering the transportation planning process in the face of bounded rationality and unbounded uncertainty Transportation Research Part C: Emerging Technologies 2005 13 2 77 92 10.1016/j.trc.2005.04.003 2-s2.0-19544389975 Gifford J. L. Checherita C. Bounded rationality and transportation behavior: lessons for public policy Transportation Research Board 86th Annual Meeting 07-2451, 2007 Cantillo V. Heydecker B. de Dios Ortúzar J. A discrete choice model incorporating thresholds for perception in attribute values Transportation Research Part B: Methodological 2006 40 9 807 825 10.1016/j.trb.2005.11.002 2-s2.0-33745173488 Cantillo V. De Dios Ortúzar J. Williams H. C. W. L. Modeling discrete choices in the presence of inertia and serial correlation Transportation Science 2007 41 2 195 205 10.1287/trsc.1060.0178 2-s2.0-67650350587 Marsden G. Frick K. T. May A. D. Deakin E. Bounded rationality in policy learning amongst cities: lessons from the transport sector Environment and Planning A 2012 44 4 905 920 10.1068/a44210 2-s2.0-84864264523 Szeto W. Y. Lo H. K. Dynamic traffic assignment: properties and extensions Transportmetrica 2006 2 1 31 52 10.1080/18128600608685654 Lou Y. Yin Y. Lawphongpanich S. Robust congestion pricing under boundedly rational user equilibrium Transportation Research Part B: Methodological 2010 44 1 15 28 10.1016/j.trb.2009.06.004 2-s2.0-71549146243 Guo X. Liu H. X. Bounded rationality and irreversible network change Transportation Research Part B: Methodological 2011 45 10 1606 1618 10.1016/j.trb.2011.05.026 2-s2.0-80455155153 Guo X. Toll sequence operation to realize target flow pattern under bounded rationality Transportation Research Part B: Methodological 2013 56 203 216 10.1016/j.trb.2013.08.006 2-s2.0-84883514290 Di X. Liu H. X. Pang J.-S. Ban X. J. Boundedly rational user equilibria (BRUE): mathematical formulation and solution sets Transportation Research Part B: Methodological 2013 57 300 313 10.1016/j.trb.2013.06.008 2-s2.0-84887822680 Di X. He X. Guo X. Liu H. X. Braess paradox under the boundedly rational user equilibria Transportation Research Part B: Methodological 2014 67 86 108 10.1016/j.trb.2014.04.005 2-s2.0-84901794015 Di X. Boundedly Rational User Equilibrium: Models and Applications 2014 University of Minnesota Han K. Szeto W. Y. Friesz T. L. Formulation, existence, and computation of boundedly rational dynamic user equilibrium with fixed or endogenous user tolerance Transportation Research Part B: Methodological 2015 79 16 49 10.1016/j.trb.2015.05.002 2-s2.0-84930938271 Mahmassani H. S. Jou R.-C. Transferring insights into commuter behavior dynamics from laboratory experiments to field surveys Transportation Research Part A: Policy and Practice 2000 34 4 243 260 10.1016/S0965-8564(99)00003-8 2-s2.0-0034025308 Patriksson P. The Traffic Assignment Problem: Models and Methods 1994 Larsson T. Patriksson M. Simplicial decomposition with disaggregated representation for the traffic assignment problem Transportation Science 1992 26 1 4 17 10.1287/trsc.26.1.4 2-s2.0-0026822031 Jayakrishnan R. Tsai W. T. Prashker J. N. Rajadhyaksha S. A Faster Path-based Algorithm for Traffic Assignment 1994 University of California Transportation Center Bar-Gera H. Origin-Based Algorithms for Transportation Network Modeling 1999 Rosenthal E. GAMS-A User's Guide 2008 GAMS Development Corporation