A robust optimization method and system for traffic user equilibrium based on uncertain supply and demand

By introducing uncertain parameters and mathematical optimization models, the problem of difficulty in getting close to real traffic situations in traditional traffic network analysis is solved, more accurate traffic flow distribution is achieved, the robustness and authenticity of the traffic model are improved, and the urban traffic is for The planning provides a more reasonable reference.

CN114925523BActive Publication Date: 2024-11-08GUANGZHOU YANG CHENG TONG CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210552406.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-05-20
Publication Date
2024-11-08
Estimated Expiration
2042-05-20

AI Technical Summary

Technical Problem

The existing traffic network analysis model is difficult to accurately approach the real road traffic conditions, especially under the influence of uncertain factors in traffic supply and demand, the traditional user equilibrium model cannot effectively simulate travelers' perception and path selection behavior of real-time traffic conditions.

Method used

Introduce uncertain road resistance function and traffic demand function to build an uncertain user equilibrium model. By minimizing the mathematical optimization model of uncertain set expectations, use the second-order cone algorithm and concave-convex conversion algorithm to convert constraints to convex constraints, and find solutions Distribution of traffic flows that are more in line with the real situation.

Benefits of technology

This method improves the robustness and authenticity of the traffic model, can more accurately reflect the complex situation of urban traffic, and provide a more reasonable reference for traffic planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114925523B_ABST
    Figure CN114925523B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of urban traffic planning and management, and discloses a robust optimization method and system for traffic user equilibrium based on uncertain supply and demand. The present invention introduces uncertain supply and demand parameters during modeling, constructs an uncertain user equilibrium model, transforms the uncertain user equilibrium model into an uncertain complementary problem, takes minimizing the expectation of the uncertain set as the objective function, takes the probability that the value of the uncertain complementary problem is not less than 0 and is greater than the probability threshold as the constraint condition, and obtains a mathematical optimization model; uses a second-order cone algorithm to transform the constraint condition of the mathematical optimization model into a convex constraint, and transforms the objective function of the mathematical optimization model into a convex function through a concave-convex transformation algorithm, and finally solves the transformed mathematical optimization model. The present invention takes the robustness of actual traffic demand and supply into the traffic model, so that the model is more in line with the actual road traffic conditions, and proposes a new solution algorithm, which greatly reduces the difficulty of model solution.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of urban traffic planning and management, and in particular to a method and system for robust optimization of traffic user equilibrium based on uncertain supply and demand. Background Art

[0002] With the acceleration of urbanization, urban bus networks are facing increasing pressure. Research on how to improve traffic efficiency, optimize bus network layout, and upgrade transportation services has become a focus of the industry.

[0003] In the classic traffic model, traffic network analysis is generally conducted from three levels: road segment, path and OD (origin-destination). Accordingly, the traffic network analysis process is accompanied by the estimation of road segment flow, path flow and OD flow. The initial approach is to assume that each traveler can find the path with the shortest travel time between the OD pairs of the corresponding trip, which constitutes the user equilibrium (UE for short). The user equilibrium model is one of the most basic models in traffic research. Its purpose is to provide decision makers with a basis for traffic planning and management through the analysis of stable traffic flow. The so-called user equilibrium model is to determine the equilibrium traffic flow and various network performance indicators based on the path selection criterion of minimizing path cost under the given traffic demand and road segment cost function. The path selection model is embedded in the user equilibrium model to simulate people's path selection behavior, and the congestion on the network is simulated by the road resistance function of the road segment. But in real life, each traveler cannot accurately know the real-time traffic conditions of the road network and accurately perceive the actual travel time. In this case, there is a deviation between the perceived time and the actual time. Using the statistical assumption of this deviation, there is a stochastic user equilibrium (SUE). However, traditional user equilibrium models, such as Wordrop equilibrium and stochastic user equilibrium, assume that traffic demand and supply are certain, which is inconsistent with the actual traffic situation.

[0004] In the actual environment, both traffic supply and demand are affected by various uncertain factors, such as traffic accidents, bad weather, traffic control, etc. Therefore, the recent research on user equilibrium models with uncertain factors has attracted more and more attention. Urban traffic travelers are the objects of traffic network analysis. In the context of increasingly rich data resources, how to better understand human travel activities is facing challenges. At the same time, under different road, traffic, control and environmental conditions, simply regressing traffic phenomena or using empirical formulas to approximate reality cannot understand traffic in essence and is difficult to get close to the real road traffic situation. Summary of the invention

[0005] The present invention provides a method and system for robust optimization of traffic user equilibrium based on uncertain supply and demand, which solves the technical problem that the user equilibrium model in the existing traffic network analysis is difficult to be close to the actual road traffic situation.

[0006] A first aspect of the present invention provides a robust optimization method for traffic user equilibrium based on uncertain supply and demand, comprising:

[0007] Assuming that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to the normal distribution, the corresponding uncertain parameters are introduced into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function;

[0008] Based on the uncertain road resistance function, an objective function of an uncertain user equilibrium model is constructed, and a traffic non-negative constraint and a traffic flow conservation constraint are used as constraint conditions of the uncertain user equilibrium model to obtain the uncertain user equilibrium model;

[0009] Based on the uncertain traffic demand function, the uncertain user equilibrium model is transformed into an uncertain complementarity problem, minimizing the uncertainty set expectation is used as the objective function, and the probability that the value of the uncertain complementarity problem is not less than 0 is greater than a probability threshold is used as a constraint condition, so as to obtain a corresponding mathematical optimization model;

[0010] The constraint conditions of the mathematical optimization model are converted into convex constraints by using a second-order cone algorithm, and the objective function of the mathematical optimization model is converted into a convex function by using a concave-convex conversion algorithm to obtain a processed mathematical optimization model, and the processed mathematical optimization model is solved.

[0011] According to an achievable method of the first aspect of the present invention, the introducing corresponding uncertain parameters into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function includes:

[0012] The uncertainty road resistance function of the constructed road section is:

[0013] Ta (v a ,ε1)=t a (v a )+r a ε1

[0014] Where, T a (v a ,ε1) represents the uncertainty road resistance of section a, v a is the traffic flow of section a, a∈A, A is the set of sections in the traffic network, t a (v a ) is the deterministic road resistance of section a, r a is the fluctuation upper bound of the impedance time on section a, ε1 is the set C consisting of the uncertain parameters of A sections 1×A A 1×A ​​vector in ;

[0015] The uncertainty resistance function of the constructed path is:

[0016]

[0017] In the formula, represents the uncertainty path resistance of path p between OD pairs w, is the relationship coefficient between section a and path p. If section a belongs to path p, then If segment a does not belong to path p, then is the traffic flow of path p between OD pair w;

[0018] The uncertain traffic demand function of OD pairs is constructed as:

[0019]

[0020] In the formula, q w is the uncertain traffic demand of OD for w, is the average traffic demand of OD for w, β w is the upper bound of the demand fluctuation of OD for w, ε2 is the set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

[0021] According to an achievable manner of the first aspect of the present invention, the step of converting the uncertain user equilibrium model into an uncertain complementarity problem based on the uncertain traffic demand function includes:

[0022] The uncertain user equilibrium model is transformed into the following uncertain complementarity problem:

[0023]

[0024] In the formula, F(x,u,ε) represents the uncertainty complementarity problem, x represents the road resistance function cost, u represents the minimum travel time of the OD pair, and ε represents the uncertain parameter. f is the path traffic flow, M is the number of paths, W is the number of OD pairs, △ is the link-path relationship matrix, C A×M represents an A×M matrix, D, and B are both diagonal matrices, and the matrix vector of D represents the travel time of section a in free flow state, t 0 yes The vector composed of 0 ∈C A×1 , C A×1 represents an A×1 matrix, c a is the capacity of section a, α is the coefficient of the US Federal Highway Administration model, The matrix vector The matrix vector B of B ww= β w , Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents the W × M matrix, q is the uncertain traffic demand of OD pairs;

[0025] By splitting the expression of F(x,u,ε), we can get the following form of uncertainty complementarity problem:

[0026]

[0027] Where C0 represents the zero matrix;

[0028] The uncertainty complementarity problem in the above form is simplified to the following form:

[0029] F(x,u,ε)=Zx+Hε+d

[0030] In the formula, Z, H and d represent a matrix respectively, where

[0031] According to an achievable manner of the first aspect of the present invention, the objective function is to minimize the uncertainty set expectation, and the constraint condition is that the probability that the value of the uncertainty complementarity problem is not less than 0 is greater than the probability threshold, to obtain a corresponding mathematical optimization model, including:

[0032] The corresponding mathematical optimization model is constructed as follows:

[0033]

[0034] In the formula, Eε represents the uncertainty set expectation, prob ε {Zx+Hε+d≥0} represents the probability that the value of the uncertainty complementarity problem is not less than 0, P is the probability threshold, and P∈(0.5,1).

[0035] According to an achievable manner of the first aspect of the present invention, the use of a second-order cone algorithm to convert the constraint conditions of the mathematical optimization model into convex constraints includes:

[0036] The constraint condition prob{Zx+Hε+d≥0}≥P is changed to:

[0037]

[0038] Assume that ε has a mathematical expectation of μ and a variance of θ 2 The normal distribution of , transforms the above formula into the following second-order cone constraint:

[0039]

[0040] In the formula, Φ represents the symbol of the standard normal distribution function, which converts the normal distribution into the standard normal distribution so as to query the value through the standard normal distribution table;

[0041] Introduce h i :

[0042] h i =Φ -1 (P)‖-θH i ‖-μ T H i

[0043] Then the equivalent form of the mathematical optimization model is as follows:

[0044]

[0045] The above formula is written in the form of a matrix, and the mathematical optimization model after the constraint conditions are transformed into convex constraints is obtained:

[0046]

[0047] In the formula,

[0048] According to an achievable manner of the first aspect of the present invention, converting the objective function of the mathematical optimization model into a convex function by a concave-convex conversion algorithm comprises:

[0049] For x in the objective function of the mathematical optimization model T Zx performs transformation:

[0050]

[0051] Through the transformation of the above formula, is a symmetric matrix. Perform eigenvalue decomposition in the following form:

[0052]

[0053] In the formula, and The elements of are all numbers greater than or equal to zero;

[0054] Will This breaks down into:

[0055]

[0056] The mathematical optimization model after the constraints are converted into convex constraints is converted into the following form:

[0057]

[0058] right Performing a second-order Taylor expansion, we get:

[0059]

[0060] The mathematical optimization model is finally transformed into the following form:

[0061]

[0062] According to an achievable manner of the first aspect of the present invention, solving the processed mathematical optimization model includes:

[0063] The processed mathematical optimization model is solved by the CVX toolbox of Matlab software.

[0064] A second aspect of the present invention provides a traffic user equilibrium robust optimization system based on uncertain supply and demand, comprising:

[0065] The uncertain parameter introduction module is used to assume that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to the normal distribution, and introduce the corresponding uncertain parameters into the road resistance function and traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function;

[0066] An uncertain user equilibrium model construction module is used to construct an objective function of the uncertain user equilibrium model based on the uncertain road resistance function, and to obtain the uncertain user equilibrium model by taking a non-negative traffic constraint and a traffic flow conservation constraint as constraint conditions of the uncertain user equilibrium model;

[0067] An optimization model determination module is used to transform the uncertain user equilibrium model into an uncertain complementary problem based on the uncertain traffic demand function, taking minimization of the uncertain set expectation as the objective function and taking the probability that the value of the uncertain complementary problem is not less than 0 greater than a probability threshold as a constraint condition, to obtain a corresponding mathematical optimization model;

[0068] The model optimization solving module is used to transform the constraint conditions of the mathematical optimization model into convex constraints by using a second-order cone algorithm, and transform the objective function of the mathematical optimization model into a convex function by using a concave-convex transformation algorithm to obtain a processed mathematical optimization model, and solve the processed mathematical optimization model.

[0069] According to an achievable manner of the second aspect of the present invention, the uncertain parameter introduction module includes:

[0070] The first function construction unit is used to construct the uncertainty road resistance function of the road section:

[0071] T a (v a ,ε1)=t a (v a )+r a ε1

[0072] Where, T a (v a ,ε1) represents the uncertainty road resistance of section a, v a is the traffic flow of section a, a∈A, A is the set of sections in the traffic network, t a (v a ) is the deterministic road resistance of section a, r a is the fluctuation upper bound of the impedance time on section a, ε1 is the set C consisting of the uncertain parameters of A sections 1×A A 1×A ​​vector in ;

[0073] The second function construction unit is used to construct the uncertainty road resistance function of the path:

[0074]

[0075] In the formula, represents the uncertainty path resistance of path p between OD pairs w, is the relationship coefficient between section a and path p. If section a belongs to path p, then If segment a does not belong to path p, then is the traffic flow of path p between OD pair w;

[0076] The third function construction unit is used to construct the uncertainty traffic demand function of the OD pair:

[0077]

[0078] In the formula, q w is the uncertain traffic demand of OD for w, is the average traffic demand of OD for w, β w is the upper bound of the demand fluctuation of OD for w, ε2 is the set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

[0079] According to an achievable manner of the second aspect of the present invention, the optimization model determination module includes a model conversion unit, and the model conversion unit is specifically used to:

[0080] The uncertain user equilibrium model is transformed into the following uncertain complementarity problem:

[0081]

[0082] In the formula, F(x,u,ε) represents the uncertainty complementarity problem, x represents the road resistance function cost, u represents the minimum travel time of the OD pair, and ε represents the uncertain parameter. f is the path traffic flow, M is the number of paths, W is the number of OD pairs, △ is the link-path relationship matrix, C A×M represents an A×M matrix, D, and B are both diagonal matrices, and the matrix vector of D represents the travel time of section a in free flow state, t 0 yes The vector composed of 0 ∈C A×1 , C A×1 represents an A×1 matrix, c a is the capacity of section a, α is the coefficient of the US Federal Highway Administration model, The matrix vector The matrix vector B of B ww= β w , Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents the W × M matrix, q is the uncertain traffic demand of OD pairs;

[0083] By splitting the expression of F(x,u,ε), we can get the following form of uncertainty complementarity problem:

[0084]

[0085] Where C0 represents the zero matrix;

[0086] The uncertainty complementarity problem in the above form is simplified to the following form:

[0087] F(x,u,ε)=Zx+Hε+d

[0088] In the formula, Z, H and d represent a matrix respectively, where

[0089] According to an achievable manner of the second aspect of the present invention, the optimization model determination module includes an optimization model construction unit, and the optimization model construction unit is specifically used to:

[0090] The corresponding mathematical optimization model is constructed as follows:

[0091]

[0092] In the formula, E ε represents the uncertainty set expectation, prob ε {Zx+Hε+d≥0} represents the probability that the value of the uncertainty complementarity problem is not less than 0, P is the probability threshold, and P∈(0.5,1).

[0093] According to an achievable manner of the second aspect of the present invention, the model optimization solution module includes a constraint optimization unit, and the constraint optimization unit is specifically used to:

[0094] The constraint condition prob{Zx+Hε+d≥0}≥P is changed to:

[0095]

[0096] Assume that ε has a mathematical expectation of μ and a variance of θ 2 The normal distribution of , transforms the above formula into the following second-order cone constraint:

[0097]

[0098] In the formula, Φ represents the symbol of the standard normal distribution function, which converts the normal distribution into the standard normal distribution so as to query the value through the standard normal distribution table;

[0099] Introduce h i :

[0100] h i =Φ -1 (P)‖-θH i ‖-μ T H i

[0101] Then the equivalent form of the mathematical optimization model is as follows:

[0102]

[0103] The above formula is written in the form of a matrix, and the mathematical optimization model after the constraint conditions are transformed into convex constraints is obtained:

[0104]

[0105] In the formula,

[0106] According to an achievable manner of the second aspect of the present invention, the model optimization solution module includes an objective function optimization unit, and the objective function optimization unit is specifically used to:

[0107] For x in the objective function of the mathematical optimization model T Zx performs transformation:

[0108]

[0109] Through the transformation of the above formula, is a symmetric matrix. Perform eigenvalue decomposition in the following form:

[0110]

[0111] In the formula, and The elements of are all numbers greater than or equal to zero;

[0112] Will This breaks down into:

[0113]

[0114] The mathematical optimization model after the constraints are converted into convex constraints is converted into the following form:

[0115]

[0116] right Performing a second-order Taylor expansion, we get:

[0117]

[0118] The mathematical optimization model is finally transformed into the following form:

[0119]

[0120] According to an achievable manner of the second aspect of the present invention, the model optimization solution module includes a model solution unit, and the model solution unit is specifically used to:

[0121] The processed mathematical optimization model is solved by the CVX toolbox of Matlab software.

[0122] The third aspect of the present invention provides a robust optimization device for traffic user equilibrium based on uncertain supply and demand, comprising:

[0123] A memory for storing instructions; wherein the instructions are used to implement the robust optimization method for traffic user equilibrium based on uncertain supply and demand as described in any of the above implementation methods;

[0124] A processor is used to execute instructions in the memory.

[0125] A fourth aspect of the present invention is a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the robust optimization method for traffic user equilibrium based on uncertain supply and demand as described in any of the above achievable methods.

[0126] It can be seen from the above technical solutions that the present invention has the following advantages:

[0127] The present invention assumes that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to the normal distribution, introduces corresponding uncertain parameters into the road resistance function and the traffic demand function respectively, and obtains the uncertain road resistance function and the uncertain traffic demand function; constructs the objective function of the uncertain user equilibrium model based on the uncertain road resistance function, takes the non-negative flow constraint and the traffic flow conservation constraint as the constraint conditions of the model, and obtains the uncertain user equilibrium model; based on the uncertain traffic demand function, transforms the uncertain user equilibrium model into an uncertain complementary problem, takes minimizing the expectation of the uncertain set as the objective function, takes the probability that the value of the uncertain complementary problem is not less than 0 is greater than the probability threshold as the constraint condition, and obtains the corresponding mathematical optimization model; uses the second-order cone algorithm to transform the constraint conditions of the mathematical optimization model into convex constraints, and through the concave-convex transformation The algorithm converts the objective function of the mathematical optimization model into a convex function, obtains the processed mathematical optimization model, and solves the processed mathematical optimization model; compared with previous traffic modeling, the present invention takes the robustness of actual traffic demand and supply into consideration in the traffic model, which is more in line with the actual road traffic conditions, and the obtained results are more realistic; the present invention provides a new model solving algorithm, which converts the constraint conditions into a second-order cone constraint by accurately convexly representing and internally convexly approximating the constraint conditions, so as to make it a convex constraint condition, and uses a convex-concave conversion algorithm on the objective function, so as to make it a convex problem to approximate and solve it, which greatly reduces the difficulty of model solving; the model solving result obtained by the present invention is more in line with the actual situation, can provide a more reasonable reference for urban traffic planning departments, and make contributions to the development of urban traffic. BRIEF DESCRIPTION OF THE DRAWINGS

[0128] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0129] Figure 1 A flowchart of a robust optimization method for traffic user equilibrium based on uncertain supply and demand provided by an optional embodiment of the present invention;

[0130] Figure 2 A principle block diagram of a traffic user equilibrium robust optimization system based on uncertain supply and demand is provided for an optional embodiment of the present invention.

[0131] Reference numerals:

[0132] 1- Uncertain parameter introduction module; 2- Uncertain user equilibrium model construction module; 3- Optimization model determination module; 4- Model optimization solution module. DETAILED DESCRIPTION

[0133] The embodiment of the present invention provides a method and system for robust optimization of traffic user equilibrium based on uncertain supply and demand, which is used to solve the technical problem that the user equilibrium model in the existing traffic network analysis is difficult to be close to the actual road traffic situation.

[0134] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0135] The invention provides a traffic user equilibrium robust optimization method based on uncertain supply and demand.

[0136] The traffic system targeted by the embodiment of the present invention is a strongly connected network G=(N, A), where N is a set of traffic network nodes and A is a set of traffic network sections.

[0137] See also Figure 1 , Figure 1 A flow chart of a robust optimization method for traffic user equilibrium based on uncertain supply and demand provided by an embodiment of the present invention is shown.

[0138] An embodiment of the present invention provides a traffic user equilibrium robust optimization method based on uncertain supply and demand, including steps S1-S4.

[0139] Step S1, assuming that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to the normal distribution, introduce corresponding uncertain parameters into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function.

[0140] The traffic modeling in the prior art is based on the assumption that the traffic supply and traffic demand of the traffic network are certain, and the traffic supply and traffic demand do not change with time. However, this is inconsistent with the real traffic environment. In this embodiment, it is assumed that the road resistance function and traffic demand of the traffic network are uncertain, and the uncertain parts are functions that conform to the normal distribution.

[0141] In one achievable manner, corresponding uncertain parameters are introduced into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function, including:

[0142] The uncertainty road resistance function of the constructed road section is:

[0143] T a (v a ,ε1)=t a (v a )+r a ε1

[0144] Where, T a (v a ,ε1) represents the uncertainty road resistance of section a, v a is the traffic flow of section a, a∈A, A is the set of sections in the traffic network, t a (v a ) is the deterministic road resistance of section a, r a is the fluctuation upper bound of the impedance time on section a, ε1 is the set C consisting of the uncertain parameters of A sections 1×A A 1×A ​​vector in ;

[0145] The uncertainty resistance function of the constructed path is:

[0146]

[0147] In the formula, represents the uncertainty path resistance of path p between OD pairs w, is the relationship coefficient between section a and path p. If section a belongs to path p, then If segment a does not belong to path p, then is the traffic flow of path p between OD pair w;

[0148] The uncertain traffic demand function of OD pairs is constructed as:

[0149]

[0150] In the formula, q w is the uncertain traffic demand of OD for w, is the average traffic demand of OD for w, β w is the upper bound of the demand fluctuation of OD for w, ε2 is the set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

[0151] According to the idea of ​​robust optimization, in the embodiment of the present invention, an assumption is made on the uncertainty set, and uncertain parameters ε1 and ε2 are introduced, where ε1 is a set C consisting of uncertain parameters of A sections. 1×A A 1×A ​​vector in , ε2 is a set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

[0152] In the embodiment of the present application, it is assumed that the uncertain parameter ε is a set belonging to a normal distribution, U = {ε∈(μ,Ω)}, μ is the mean vector of the uncertain parameter, and Ω is the variance matrix of the uncertain set. In the subsequent steps, the present application assumes that:

[0153] The uncertain road resistance function of each path and the uncertain traffic demand of each OD pair are bounded for any ε∈U; the uncertain road resistance function of each path is a non-negative smooth function.

[0154] The road resistance function should reflect the law of traffic congestion, which usually requires it to be nonlinear and strictly increase with the flow rate. The most classic road resistance function expression is the standard BPR (Bureau of Public Road, Federal Highway Administration of the United States) model, that is,

[0155]

[0156] In the formula, represents the travel time of road section a in free flow state (vehicles can travel "freely" in a state with very little traffic volume), c a refers to the traffic capacity of section a, v a The flow rate on the road section a∈A. In addition, α and τ in the BPR model are constants and can be changed.

[0157] In the embodiment of the present invention, the deterministic road resistance function of each road section is a model function of the Federal Highway Administration of the United States. It should be noted that the deterministic road resistance function of each road section may also be a road resistance function in other forms.

[0158] In the above embodiment of the present invention, the robustness of actual traffic demand and supply is taken into account in the traffic model, and specific expressions of uncertain road resistance function and uncertain traffic demand function are given, which provides a functional basis for the subsequent construction of uncertain user equilibrium model and corresponding mathematical optimization model.

[0159] Step S2, constructing the objective function of the uncertain user equilibrium model based on the uncertain road resistance function, taking the traffic non-negative constraint and the traffic flow conservation constraint as the constraint conditions of the uncertain user equilibrium model, and obtaining the uncertain user equilibrium model.

[0160] According to the user equilibrium principle proposed by Wardrop, when the network is in equilibrium, all travelers in the traffic network cannot reduce the impedance time by unilaterally changing the path. At this time, if the Wardrop user equilibrium criterion is directly expressed in mathematical language, it can be described as:

[0161] When the traffic network reaches equilibrium, if there is Must have is the path p of OD to w in the traffic network, is the relationship coefficient between the path p of the section a and the OD pair w, u w is the minimum travel time of OD to w. This means that if two or more paths are selected for OD to w, their travel times are equal. Must have This means that if the path flow of a certain OD pair is equal to zero, then the travel time of this path must exceed the travel time of the selected path.

[0162] According to the user equilibrium principle, the following constraints should be met during the balanced allocation process:

[0163] There is an identity constraint on the flow between the path and the OD, that is, the sum of the traffic flow on each path between the OD pairs should be equal to the total OD traffic flow. It can be expressed as:

[0164]

[0165] In the formula, represents the traffic flow of path k of OD pair (r,s), q rs represents the total traffic flow of OD pair (r, s), w rs Represents the set of paths in the OD pair (r,s).

[0166] Secondly, there should be a certain relationship between the path traffic flow and the section traffic volume, that is, the traffic flow on the section should be the accumulation of the flow of each OD pair through the path of the section. The specific formula is as follows:

[0167]

[0168] In the formula, x a is the traffic flow of section a, is the relationship coefficient between the road segment a and the path k in the OD pair (r, s), R is the set of starting points of the transportation network, and S is the set of end points of the transportation network;

[0169] At the same time, the total impedance of the path and the impedance of the section should satisfy the following constraints, that is, the impedance of the path should be the accumulation of the impedances of the various sections along which the path passes, and the formula is expressed as:

[0170]

[0171] In the formula, is the total resistance of path k in the OD pair (r, s).

[0172] Finally, the flow of the path should satisfy the non-negative constraint, that is,

[0173] In the embodiment of the present invention, the above constraints are used as constraint conditions of the uncertain user equilibrium model.

[0174] Step S3, based on the uncertain traffic demand function, convert the uncertain user equilibrium model into an uncertain complementary problem, take minimizing the uncertainty set expectation as the objective function, take the probability that the value of the uncertain complementary problem is not less than 0 and is greater than the probability threshold as the constraint condition, and obtain the corresponding mathematical optimization model.

[0175] When modeling based on unchanged traffic demand and traffic supply, according to the user equilibrium principle proposed by Wardrop, when the network is in equilibrium, all travelers in the traffic network cannot reduce the road block time by unilaterally changing the route. Therefore, the user equilibrium deterministic model can be written as the following deterministic complementarity problem (CP):

[0176]

[0177]

[0178] x T F(x,u)=0

[0179] Where F(x,u) represents the deterministic complementarity problem, f is the path traffic flow, M is the number of paths, W is the number of OD pairs, Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents a W×M matrix, u represents the minimum travel time of the OD pair, x represents the road resistance function cost, represents the road resistance function of the path, and Q is the vector composed of the deterministic traffic demands of OD pairs.

[0180] The complementarity problem states that in equilibrium, when the flow on path p Then F(x,u) = 0. That is, the impedance function cost on the path with non-zero flow is the smallest, and the impedance time on the path with zero flow is greater than the minimum impedance time of the OD. The demand for OD pairs on the transportation network is equal to the sum of the traffic flows of all paths of the OD pair.

[0181] Based on the above-mentioned deterministic complementarity problem, in one achievable manner, based on the uncertain traffic demand function, the uncertain user equilibrium model is transformed into the following uncertain complementarity problem:

[0182]

[0183] In the formula, F(x,u,ε) represents the uncertainty complementarity problem, x represents the road resistance function cost, u represents the minimum travel time of the OD pair, and ε represents the uncertain parameter. f is the path traffic flow, M is the number of paths, W is the number of OD pairs, △ is the link-path relationship matrix, C A×M represents an A×M matrix, D, and B are both diagonal matrices, and the matrix vector of D represents the travel time of section a in free flow state, t 0 yes The vector composed of 0 ∈C A×1 , C A×1 represents an A×1 matrix, c a is the capacity of section a, α is the coefficient of the US Federal Highway Administration model, The matrix vector The matrix vector B of B ww= β w , Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents the W × M matrix, q is the uncertain traffic demand of OD pairs;

[0184] By splitting the expression of F(x,u,ε), we can get the following form of uncertainty complementarity problem:

[0185]

[0186] Where C0 represents the zero matrix;

[0187] The uncertainty complementarity problem in the above form is simplified to the following form:

[0188] F(x,u,ε)=Zx+Hε+d

[0189] In the formula, Z, H and d represent a matrix respectively, where

[0190] In an achievable manner, the objective function is to minimize the uncertainty set expectation, and the constraint condition is that the probability that the value of the uncertainty complementarity problem is not less than 0 is greater than the probability threshold, to obtain a corresponding mathematical optimization model, including:

[0191] The corresponding mathematical optimization model is constructed as follows:

[0192]

[0193] In the formula, E ε represents the uncertainty set expectation, prob ε {Zx+Hε+d≥0} represents the probability that the value of the uncertainty complementarity problem is not less than 0, P is the probability threshold, and P∈(0.5,1).

[0194] In a real traffic environment, due to the complexity of the traffic network, the traffic network is changing all the time, and F(x,u,ε) does not necessarily satisfy the condition of being greater than or equal to zero. Since the minimum impedance time function is a constant assumed by experience, the road resistance function is not necessarily greater than the minimum impedance time function, and the traffic demand is also a constant assumed by experience, so the traffic supply is not necessarily greater than the traffic demand. Therefore, according to the complementary constraints of user equilibrium, the real traffic environment considered, and the definition of robust optimization, an embodiment of the present invention constructs a minimized uncertain expectation, and the constraint is the probability that F(x,u,ε) is not less than 0, and this probability must be greater than a relatively large value. Therefore, in this embodiment, the probability threshold is set to a value between 0.5 and 1.

[0195] Step S4, using a second-order cone algorithm to transform the constraint conditions of the mathematical optimization model into convex constraints, and transforming the objective function of the mathematical optimization model into a convex function through a concave-convex transformation algorithm to obtain a processed mathematical optimization model, and solving the processed mathematical optimization model.

[0196] For the mathematical optimization model of the above steps, its objective function is to minimize the expectation of the uncertain set, and the constraint is that the probability constraint is greater than P, and P is a constant greater than 0.5. For the mathematical optimization model of the above steps, it can be seen that the solution problem of the mathematical optimization model of the above steps is not a convex optimization problem. For the objective function, since it is unknown whether Z is a positive definite matrix, if it is a negative objective function, it is a non-convex problem. For the constraint, the constraint is a probabilistic problem and also a non-convex problem, so the whole problem is a non-convex problem and is difficult to solve.

[0197] In order to solve the mathematical optimization model of the above steps, in the following embodiment of the present invention, the constraint conditions of the mathematical optimization model are converted into convex constraints based on Theorem 1.

[0198] Theorem 1: Assume that the disturbance variable ε in the probability constraint conforms to a Gaussian distribution with a variance greater than zero. If the probability is greater than 0.5, then this probability constraint is equivalent to a second-order cone constraint.

[0199] Considering linear constraints

[0200] a T x≤b

[0201] Among them, a~N(μ,θ 2 ) is a Gaussian random vector;

[0202] The probability constraint is:

[0203] prob{a T x≤b}≥p

[0204] When p≥0.5, it is equivalent to the second-order cone constraint: Φ -1 (p)‖θx‖≤b-μ T x.

[0205] Based on Theorem 1, in one achievable manner, the use of a second-order cone algorithm to convert the constraints of the mathematical optimization model into convex constraints includes:

[0206] The constraint condition prob{Zx+Hε+d≥0}≥P is changed to:

[0207]

[0208] Assume that ε has a mathematical expectation of μ and a variance of θ 2 The normal distribution of , transforms the above formula into the following second-order cone constraint:

[0209]

[0210] In the formula, Φ represents the symbol of the standard normal distribution function, which converts the normal distribution into the standard normal distribution so as to query the value through the standard normal distribution table;

[0211] Introduce h i :

[0212] h i =Φ -1 (P)‖-θH i ‖-μ T H i

[0213] Then the equivalent form of the mathematical optimization model is as follows:

[0214]

[0215] The above formula is written in the form of a matrix, and the mathematical optimization model after the constraint conditions are transformed into convex constraints is obtained:

[0216]

[0217] In the formula,

[0218] In the above embodiment of the present invention, the constraint condition is converted into a convex second-order cone constraint through the convex representation of the Gaussian perturbation vector probability constraint.

[0219] For the mathematical optimization model after the constraint conditions are converted into convex constraints, its constraints are all convex constraints, but its objective function is still a non-convex function, so the solution of the mathematical optimization model after the constraint conditions are converted into convex constraints is a non-convex optimization problem. For the objective function, in this embodiment, approximation is performed by a concave-convex conversion algorithm.

[0220] Specifically, the objective function of the mathematical optimization model is converted into a convex function by using a concave-convex conversion algorithm, including:

[0221] For x in the objective function of the mathematical optimization model T Zx performs transformation:

[0222]

[0223] Through the transformation of the above formula, is a symmetric matrix. Perform eigenvalue decomposition in the following form:

[0224]

[0225] In the formula, and The elements of are all numbers greater than or equal to zero;

[0226] Will This breaks down into:

[0227]

[0228] The mathematical optimization model after the constraints are converted into convex constraints is converted into the following form:

[0229]

[0230] For the mathematical optimization model listed above, the objective function is a non-convex function. Analysis of the objective function shows that the objective function is a non-convex function consisting of a convex function and a concave function. In this embodiment, the objective function is approximated by a concave-convex conversion algorithm.

[0231] In this embodiment, further Performing a second-order Taylor expansion, we get:

[0232]

[0233] Finally, the mathematical optimization model is transformed into the following form:

[0234]

[0235] For the above mathematical optimization model, the objective function and constraints are both convex functions, so the solution of the above mathematical optimization model is a convex problem and can be solved directly.

[0236] In one achievable manner, solving the processed mathematical optimization model includes:

[0237] The processed mathematical optimization model is solved by the CVX toolbox of Matlab software.

[0238] As a specific implementation method, the specific solution process can be performed according to the approximation algorithm in Table 1.

[0239] Table 1:

[0240]

[0241]

[0242] The above embodiments of the present invention have the following beneficial effects:

[0243] 1) Compared with the previous traffic modeling, the present invention takes the robustness of actual traffic demand and supply into consideration in the traffic model, which is more consistent with the actual road traffic conditions and the obtained results are more realistic;

[0244] 2) The present invention proposes an innovative algorithm to approximate the original non-convex optimization problem, and innovatively transforms the original optimization problem that is difficult to solve into a second-order cone constraint by accurately expressing the constraint problem and converting the internal convex approximation into a convex constraint problem. The convex-concave conversion algorithm is used on the objective function to turn it into a convex problem for approximation and solution.

[0245] 3) The results obtained through the model and algorithm of the present invention are more consistent with the actual cleaning, can provide a more reasonable reference for urban traffic planning departments, and contribute to the development of urban traffic.

[0246] The present invention also provides a traffic user equilibrium robust optimization system based on uncertain supply and demand. The system can be used to execute the traffic user equilibrium robust optimization method based on uncertain supply and demand as described in any one of the above embodiments.

[0247] See also Figure 2 , Figure 2 A principle block diagram of a traffic user equilibrium robust optimization system based on uncertain supply and demand provided by an embodiment of the present invention is shown.

[0248] An embodiment of the present invention provides a traffic user equilibrium robust optimization system based on uncertain supply and demand, comprising:

[0249] Uncertain parameter introduction module 1 is used to assume that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to normal distribution, and introduce corresponding uncertain parameters into the road resistance function and traffic demand function respectively to obtain uncertain road resistance function and uncertain traffic demand function;

[0250] The uncertain user equilibrium model construction module 2 is used to construct the objective function of the uncertain user equilibrium model based on the uncertain road resistance function, and use the traffic non-negative constraint and the traffic flow conservation constraint as the constraint conditions of the uncertain user equilibrium model to obtain the uncertain user equilibrium model;

[0251] The optimization model determination module 3 is used to transform the uncertain user equilibrium model into an uncertain complementary problem based on the uncertain traffic demand function, taking minimization of the uncertain set expectation as the objective function and taking the probability that the value of the uncertain complementary problem is not less than 0 greater than a probability threshold as the constraint condition, to obtain the corresponding mathematical optimization model;

[0252] The model optimization solving module 4 is used to transform the constraint conditions of the mathematical optimization model into convex constraints by using a second-order cone algorithm, and transform the objective function of the mathematical optimization model into a convex function by using a concave-convex transformation algorithm, so as to obtain a processed mathematical optimization model and solve the processed mathematical optimization model.

[0253] In one achievable manner, the uncertain parameter introduction module 1 includes:

[0254] The first function construction unit is used to construct the uncertainty road resistance function of the road section:

[0255] T a (v a ,ε1)=t a (v a )+r a ε1

[0256] Where, T a (v a ,ε1) represents the uncertainty road resistance of section a, v a is the traffic flow of section a, a∈A, A is the set of sections in the traffic network, t a (v a ) is the deterministic road resistance of section a, r a is the fluctuation upper bound of the impedance time on section a, ε1 is the set C consisting of the uncertain parameters of A sections 1×A A 1×A ​​vector in ;

[0257] The second function construction unit is used to construct the uncertainty road resistance function of the path:

[0258]

[0259] In the formula, represents the uncertainty path resistance of path p between OD pairs w, is the relationship coefficient between section a and path p. If section a belongs to path p, then If segment a does not belong to path p, then is the traffic flow of path p between OD pair w;

[0260] The third function construction unit is used to construct the uncertainty traffic demand function of the OD pair:

[0261]

[0262] In the formula, q w is the uncertain traffic demand of OD for w, is the average traffic demand of OD for w, β w is the upper bound of the demand fluctuation of OD for w, ε2 is the set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

[0263] In one achievable manner, the optimization model determination module 3 includes a model conversion unit, and the model conversion unit is specifically used to:

[0264] The uncertain user equilibrium model is transformed into the following uncertain complementarity problem:

[0265]

[0266] In the formula, F(x,u,ε) represents the uncertainty complementarity problem, x represents the road resistance function cost, u represents the minimum travel time of the OD pair, and ε represents the uncertain parameter. f is the path traffic flow, M is the number of paths, W is the number of OD pairs, △ is the link-path relationship matrix, C A×M represents an A×M matrix, D, and B are both diagonal matrices, and the matrix vector of D represents the travel time of section a in free flow state, t 0 yes The vector composed of 0 ∈C A×1 , C A×1 represents an A×1 matrix, c a is the capacity of section a, α is the coefficient of the US Federal Highway Administration model, The matrix vector The matrix vector B of B ww= β w , Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents the W × M matrix, q is the uncertain traffic demand of OD pairs;

[0267] By splitting the expression of F(x,u,ε), we can get the following form of uncertainty complementarity problem:

[0268]

[0269] Where C0 represents the zero matrix;

[0270] The uncertainty complementarity problem in the above form is simplified to the following form:

[0271] F(x,u,ε)=Zx+Hε+d

[0272] In the formula, Z, H and d represent a matrix respectively, where

[0273] In one achievable manner, the optimization model determination module 3 includes an optimization model construction unit, and the optimization model construction unit is specifically used to:

[0274] The corresponding mathematical optimization model is constructed as follows:

[0275]

[0276] In the formula, E ε represents the uncertainty set expectation, prob ε {Zx+Hε+d≥0} represents the probability that the value of the uncertainty complementarity problem is not less than 0, P is the probability threshold, and P∈(0.5,1).

[0277] In one achievable manner, the model optimization solution module 4 includes a constraint optimization unit, and the constraint optimization unit is specifically used for:

[0278] The constraint condition prob{Zx+Hε+d≥0}≥P is changed to:

[0279]

[0280] Assume that ε has a mathematical expectation of μ and a variance of θ 2 The normal distribution of , transforms the above formula into the following second-order cone constraint:

[0281]

[0282] In the formula, Φ represents the symbol of the standard normal distribution function, which converts the normal distribution into the standard normal distribution so as to query the value through the standard normal distribution table;

[0283] Introduce h i :

[0284] h i =Φ -1 (P)‖-θH i ‖-μ T H i

[0285] Then the equivalent form of the mathematical optimization model is as follows:

[0286]

[0287] The above formula is written in the form of a matrix, and the mathematical optimization model after the constraint conditions are transformed into convex constraints is obtained:

[0288]

[0289] In the formula,

[0290] In one achievable manner, the model optimization solution module 4 includes an objective function optimization unit, and the objective function optimization unit is specifically used to:

[0291] For x in the objective function of the mathematical optimization model T Zx performs transformation:

[0292]

[0293] Through the transformation of the above formula, is a symmetric matrix. Perform eigenvalue decomposition in the following form:

[0294]

[0295] In the formula, and The elements of are all numbers greater than or equal to zero;

[0296] Will This breaks down into:

[0297]

[0298] The mathematical optimization model after the constraints are converted into convex constraints is converted into the following form:

[0299]

[0300] right Performing a second-order Taylor expansion, we get:

[0301]

[0302] The mathematical optimization model is finally transformed into the following form:

[0303]

[0304] In one achievable manner, the model optimization solution module 4 includes a model solution unit, and the model solution unit is specifically used for:

[0305] The processed mathematical optimization model is solved by the CVX toolbox of Matlab software.

[0306] The present invention also provides a traffic user equilibrium robust optimization device based on uncertain supply and demand, comprising:

[0307] A memory for storing instructions; wherein the instructions are used to implement the robust optimization method for traffic user equilibrium based on uncertain supply and demand as described in any one of the above embodiments;

[0308] A processor is used to execute instructions in the memory.

[0309] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for robust optimization of traffic user equilibrium based on uncertain supply and demand as described in any one of the above embodiments is implemented.

[0310] Technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments, and the specific beneficial effects of the systems, devices and modules described above can refer to the corresponding beneficial effects in the aforementioned method embodiments, which will not be repeated here.

[0311] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, the indirect coupling or communication connection of the modules can be electrical, mechanical or other forms.

[0312] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0313] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of software functional modules.

[0314] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0315] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A robust optimization method for traffic user equilibrium based on uncertain supply and demand, characterized in that: include: Assuming that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to normal distribution, corresponding uncertain parameters are introduced into the road resistance function and traffic demand function respectively to obtain uncertain road resistance function and uncertain traffic demand function; Based on the uncertain road resistance function, an objective function of an uncertain user equilibrium model is constructed, and a traffic non-negative constraint and a traffic flow conservation constraint are used as constraint conditions of the uncertain user equilibrium model to obtain the uncertain user equilibrium model; Based on the uncertain traffic demand function, the uncertain user equilibrium model is transformed into an uncertain complementarity problem, minimizing the uncertainty set expectation is used as the objective function, and the probability that the value of the uncertain complementarity problem is not less than 0 is greater than a probability threshold is used as a constraint condition, so as to obtain a corresponding mathematical optimization model; The constraint conditions of the mathematical optimization model are converted into convex constraints by using a second-order cone algorithm, and the objective function of the mathematical optimization model is converted into a convex function by using a concave-convex conversion algorithm to obtain a processed mathematical optimization model, and the processed mathematical optimization model is solved.

2. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 1 is characterized in that: The corresponding uncertain parameters are introduced into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function, including: The uncertainty road resistance function of the constructed road section is: T a (v a ,ε1)=t a (v a )+r a ε1 Where, T a (v a ,ε1) represents the uncertainty road resistance of section a, v a is the traffic flow of section a, a∈A, A is the set of sections in the traffic network, t a (v a ) is the deterministic road resistance of section a, r a is the fluctuation upper bound of the impedance time on section a, ε1 is the set C consisting of the uncertain parameters of A sections 1×A A 1×A ​​vector in ; The uncertainty resistance function of the constructed path is: In the formula, represents the uncertainty path resistance of path p between OD pairs w, is the relationship coefficient between section a and path p. If section a belongs to path p, then If segment a does not belong to path p, then is the traffic flow of path p between OD pair w; The uncertain traffic demand function of OD pairs is constructed as: In the formula, q w is the uncertain traffic demand of OD for w, is the average traffic demand of OD for w, β w is the upper bound of the demand fluctuation of OD for w, ε2 is the set C consisting of W OD pairs of uncertain parameters 1×W A 1×W vector in .

3. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 2 is characterized in that: The step of converting the uncertain user equilibrium model into an uncertain complementarity problem based on the uncertain traffic demand function includes: The uncertain user equilibrium model is transformed into the following uncertain complementarity problem: In the formula, F(x,u,ε) represents the uncertainty complementarity problem, x represents the road resistance function cost, u represents the minimum travel time of the OD pair, and ε represents the uncertain parameter. f is the path traffic flow, M is the number of paths, W is the number of OD pairs, △ is the link-path relationship matrix, C A×M represents an A×M matrix, D, and B are both diagonal matrices, and the matrix vector of D represents the travel time of section a in free flow state, t 0 yes The vector composed of 0 ∈C A ×1 , C A×1 represents an A×1 matrix, c a is the capacity of section a, α is the coefficient of the US Federal Highway Administration model, The matrix vector The matrix vector B of B ww =β w , Λ is the relationship matrix between OD pairs and paths, Λ∈C W×M , C W×M represents the W × M matrix, q is the uncertain traffic demand of OD pairs; By splitting the expression of F(x,u,ε), we can get the following form of uncertainty complementarity problem: Where C0 represents the zero matrix; The uncertainty complementarity problem in the above form is simplified to the following form: F(x,u,ε)=Zx+Hε+d In the formula, Z, H and d represent a matrix respectively, where 4. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 3 is characterized in that: The objective function is to minimize the expectation of the uncertain set, and the constraint condition is that the probability that the value of the uncertainty complementarity problem is not less than 0 is greater than the probability threshold, and the corresponding mathematical optimization model is obtained, including: The corresponding mathematical optimization model is constructed as follows: In the formula, E ε represents the uncertainty set expectation, prob ε {Zx+Hε+d≥0} represents the probability that the value of the uncertainty complementarity problem is not less than 0, P is the probability threshold, and P∈(0.5,1).

5. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 4 is characterized in that: The method of converting the constraint conditions of the mathematical optimization model into convex constraints by using a second-order cone algorithm includes: The constraint condition prob{Zx+Hε+d≥0}≥P is changed to: Assume that ε has a mathematical expectation of μ and a variance of θ 2 The normal distribution of , transforms the above formula into the following second-order cone constraint: In the formula, Φ represents the symbol of the standard normal distribution function, which converts the normal distribution into the standard normal distribution so as to query the value through the standard normal distribution table; Introduce h i : h i =Φ -1 (P)‖-θH i ‖-m T H i Then the equivalent form of the mathematical optimization model is as follows: The above formula is written in the form of a matrix, and the mathematical optimization model after the constraint conditions are transformed into convex constraints is obtained: In the formula, 6. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 5 is characterized in that: The step of converting the objective function of the mathematical optimization model into a convex function by a concave-convex conversion algorithm comprises: For x in the objective function of the mathematical optimization model T Zx performs transformation: Through the transformation of the above formula, is a symmetric matrix. Perform eigenvalue decomposition in the following form: In the formula, and The elements of are all numbers greater than or equal to zero; Will This breaks down into: The mathematical optimization model after the constraints are converted into convex constraints is converted into the following form: right Performing a second-order Taylor expansion, we get: The mathematical optimization model is finally transformed into the following form:

7. The method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to claim 6 is characterized in that: The step of solving the processed mathematical optimization model comprises: The processed mathematical optimization model is solved by the CVX toolbox of Matlab software.

8. A robust optimization system for traffic user equilibrium based on uncertain supply and demand, characterized in that: include: The uncertain parameter introduction module is used to assume that the road resistance function and traffic demand of the traffic network are uncertain parts and conform to the normal distribution, and introduce the corresponding uncertain parameters into the road resistance function and the traffic demand function respectively to obtain the uncertain road resistance function and the uncertain traffic demand function; An uncertain user equilibrium model construction module is used to construct an objective function of the uncertain user equilibrium model based on the uncertain road resistance function, and to obtain the uncertain user equilibrium model by taking a non-negative traffic constraint and a traffic flow conservation constraint as constraint conditions of the uncertain user equilibrium model; An optimization model determination module is used to transform the uncertain user equilibrium model into an uncertain complementary problem based on the uncertain traffic demand function, taking minimization of the uncertain set expectation as the objective function and taking the probability that the value of the uncertain complementary problem is not less than 0 greater than a probability threshold as a constraint condition, to obtain a corresponding mathematical optimization model; The model optimization solving module is used to transform the constraint conditions of the mathematical optimization model into convex constraints by using a second-order cone algorithm, and transform the objective function of the mathematical optimization model into a convex function by using a concave-convex transformation algorithm, so as to obtain a processed mathematical optimization model and solve the processed mathematical optimization model.

9. A robust optimization device for traffic user equilibrium based on uncertain supply and demand, characterized in that: include: A memory for storing instructions; wherein the instructions are used to implement the robust optimization method for traffic user equilibrium based on uncertain supply and demand according to any one of claims 1 to 7; A processor is used to execute instructions in the memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for robust optimization of traffic user equilibrium based on uncertain supply and demand according to any one of claims 1 to 7 is implemented.