A method for predicting residents' travel demand for urban shared transportation
By building a multi-mode, multi-user random ride-sharing user balanced transportation network and logit behavior selection model, we predict the travel needs of urban transportation networks, and solve the problems that are difficult to consider in the impact of multiple travel methods and service levels in the existing methods, and improve the prediction Accuracy and applicability.
Patent Information
- Application Number
- CN202110918310.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-08-11
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2041-08-11
AI Technical Summary
The existing urban transportation demand forecasting methods are difficult to effectively consider the impact of multiple travel methods and service levels, resulting in the inaccurate prediction results.
A method for predicting residents' travel demand for shared transportation in cities is proposed. By establishing the topological structure of the urban transportation network, a multi-mode, multi-user random ride-sharing user balanced transportation network is built. Combining the traveler's perceived difference in travel costs, the logit behavior selection model is adopted. Conduct traffic allocation and predict the travel service level and demand of the transportation network.
The applicability and accuracy of the prediction method are improved and the travel characteristics of modern urban residents can be better reflected. Especially in the online car-hailing travel mode, the impact mechanism of shared travel on travel needs is clarified.
Smart Images

Figure CN113570418B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of urban transportation, and in particular relates to a method for predicting resident travel demand for urban shared transportation. Background Art
[0002] The rapid advancement of Internet technology has not only brought about the vigorous development of the mobile travel industry, but also profoundly changed people's travel concepts. As the concept of "shared travel" has become popular, online car-hailing (also known as ridesharing) has gradually become a daily travel option for many urban residents. The results of the "2021 Urban Transportation Questionnaire Survey" jointly launched by the "Daily Economic News" and Tencent Auto show that private cars, public transportation and online car-hailing are the main modes of transportation for urban residents at present, and the three cover nearly 90% of travel needs. Based on the travel characteristics of modern urban residents in the context of shared travel, a method for predicting residents' travel demand suitable for modern urban transportation networks is proposed, which can provide a reliable basis for the formulation and adjustment of transportation demand management policies and shared travel enterprise business strategies. Summary of the invention
[0003] In order to solve the technical problems mentioned in the above background technology, the present invention proposes a method for predicting residents' travel demand for urban shared transportation.
[0004] In order to achieve the above technical objectives, the technical solution of the present invention is:
[0005] A method for predicting resident travel demand for urban shared transportation comprises the following steps:
[0006] (1) Establish the topological structure of the urban transportation network. The travel modes in the urban transportation network include private car travel, online car-hailing travel, and public transportation travel. The user types in the urban transportation network include car users and non-car users. Obtain the perception parameters of travelers and the operating parameters of public transportation companies and online car-hailing platforms;
[0007] (2) Obtain the functional relationship between urban residents’ travel demand and travel service level based on historical data, namely, the elastic demand function;
[0008] (3) Based on the topological structure of the urban transportation network, user types, and travel mode information, a multi-mode multi-user random carpooling user equilibrium transportation network corresponding to the urban network is constructed, and carpooling matching constraints are established based on the network;
[0009] (4) Based on the actual characteristics of each travel mode in the urban transportation network and the individual differences in travelers' perception of travel costs, a generalized travel cost applicable to the shared travel environment is constructed;
[0010] (5) Estimate the travel service level of the city's transportation network under the equilibrium state of multi-user random carpooling;
[0011] (6) Based on the elastic demand function, combined with the obtained travel service level of the transportation network and the generalized travel cost under this service level, predict the travel demand of residents in the urban transportation network.
[0012] Furthermore, in step (2), the functional relationship between the travel demand of urban residents and the travel service level is as follows:
[0013]
[0014]
[0015] In the above formula, and They represent the travel demands of residents with and without cars between w by OD; and represents the elastic demand function; and They represent the upper limits of OD's travel demand for car users and car-free users between w; μ represents the parameter in the exponential function; and They represent the travel service levels felt by car users and non-car users in w, and The larger the value, the higher the level of travel service:
[0016]
[0017]
[0018] In the above formula, i∈C and i∈NC represent car users and car-free users respectively; represents the generalized travel cost borne by the traveler who chooses travel mode i on path p connecting OD pair w; θ represents the parameter representing the traveler's perception error in the logit choice model.
[0019] Furthermore, the specific process of step (3) is as follows:
[0020] (301) Inputting the topological structure, user type and travel mode information of the urban road network, and constructing a multi-mode multi-user urban transportation network structure corresponding to the urban transportation network in the software, the transportation network structure including the connection status of nodes and road sections in the road network, and the road section capacity and length information;
[0021] In terms of user type, according to whether the travelers in the network own private cars, they are divided into car users and car-free users, represented by i∈C and i∈NC respectively;
[0022] In terms of travel modes, the three modes of online car-hailing, private car travel and bus travel in the transportation network are divided according to travel roles, resulting in four travel modes: private car drivers, online car-hailing drivers, online car-hailing passengers and bus passengers, corresponding to i∈SD, i∈RD, i∈R, i∈PT respectively; the above travel modes are further subdivided according to whether or not a person owns a private car: RD=RD c ∪RD nc , R=R c ∪R nc and PT = PT c ∪PT nc ; where i∈R c and i∈R nc They represent online ride-hailing passengers with cars and online ride-hailing passengers without cars respectively; i∈RD c Indicates that the online car-hailing passenger i∈R is carried c of online car-hailing drivers, i∈RD nc Indicates that the online car-hailing passenger i∈R is carried nc of online ride-hailing drivers; i∈PT c and i∈PT nc They represent bus passengers with cars and bus passengers without cars respectively;
[0023] (302) Based on the constructed multi-mode and multi-user urban transportation network structure, the ride-sharing matching constraints are established:
[0024]
[0025] In the above formula, represents the flow of travelers who choose mode i on path p connecting OD pair w, Γ r (i) represents the online car-hailing driver i∈RD corresponding to the online car-hailing passenger i∈R, N i represents the number of passengers in the car of online car-hailing driver i;
[0026]
[0027]
[0028] In the above formula, represents the flow of online car-hailing drivers i∈RD between OD and w, that is, the supply of online car-hailing travel services; Denotes OD for w online ride-hailing passengers Γ r (i)∈R, that is, the demand for online car-hailing services.
[0029] Furthermore, in step (4), the generalized travel cost is:
[0030]
[0031] In the above formula, Represents the travel time perceived by travelers:
[0032]
[0033] In the above formula, ρ i refers to the time value parameter of the traveler who chooses mode i, and They refer to the driving time of cars on path p and road section a respectively. and They refer to the bus travel time on path p and section a respectively, Represents the segment-path conversion coefficient. If path p contains segment a, otherwise,
[0034] Expressing inconvenience experience in the car:
[0035]
[0036] In the above formula, γ i is the inconvenience experience parameter of travelers who choose mode i;
[0037] Express the cost of travellers:
[0038]
[0039] In the above formula, shows the mileage of path p between OD pairs w, b i , and r i They are the travel time parameter, mileage parameter and floating pricing parameter set by the online car-hailing platform, τ b It represents the bus mileage parameters set by the bus company. represents the Lagrange multiplier generated by the co-multiplication matching constraint, Γ rd (i) represents matching the online car-hailing passenger i∈R to the responding online car-hailing driver;
[0040] c f represents the fixed cost of a private car, c t Represents the cost of using a private car.
[0041] Furthermore, the specific process of step (5) is as follows:
[0042] (501) Considering the randomness and differences in travelers’ perception of generalized travel costs, a logit-based travel choice model is used in the flow allocation process, and the path flow is obtained as follows:
[0043]
[0044]
[0045]
[0046] In the above formula, represents the probability that a traveler between OD pair w chooses mode i on path p;
[0047] (502) In the feasible set Ω, find the vector f * ∈Ω satisfies:
[0048]
[0049] In the above formula, represents the path flow vector in the network, W is the set of all OD pairs in the urban transportation network, P w is the set of all paths between OD pairs w, I represents the set of all travel modes; Ω represents the set of path flow vectors that satisfy the ride-sharing matching constraints. The specific expression is:
[0050]
[0051] in, The expression is:
[0052]
[0053] and Denote the demand function and The inverse function of is expressed as:
[0054]
[0055]
[0056] f * represents the solution of the variational inequality, that is, the traffic flow of the urban transportation network under the equilibrium state of multi-user random carpooling; based on this, the travel service level of the urban transportation network under the equilibrium state of multi-user random carpooling is obtained. and
[0057] Furthermore, in step (6), according to the travel service level obtained in step (5) and Combined with the elastic demand function obtained in step (2) and Predict the travel needs of residents in the city's transportation network:
[0058]
[0059]
[0060] The beneficial effects brought by adopting the above technical solution are:
[0061] ① The present invention is based on the general background of the sharing economy, with online car-hailing, private car and public transportation as the main travel modes, making the travel demand forecasting method applicable to a variety of travel modes, thereby improving the applicability of the forecasting method.
[0062] ② The present invention takes into account the impact of the service level in the transportation network on the travel needs of residents, and is based on the urban transportation network with multiple types of travelers, and uses logit behavior selection as the flow distribution principle. The travel demand prediction process is more in line with the travel characteristics of modern urban residents, making the prediction results more accurate.
[0063] ③ According to the characteristics of travel modes in the network, the present invention has carried out a refined classification of the broad travel costs, especially an in-depth exploration of the online car-hailing travel model, making the impact mechanism of "shared travel" on the travel needs of modern urban residents clearer. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] Figure 1 is a flow chart of the method of the present invention;
[0065] Figure 2 It is a schematic diagram of the urban transportation network structure;
[0066] Figure 3 It is the travel demand map of the urban transportation network. DETAILED DESCRIPTION
[0067] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.
[0068] The present invention provides a method for predicting the travel demand of residents in urban shared transportation. Figure 1 As shown, the specific process is as follows:
[0069] 1) Understand the current status of the road network in the target city or region by reading maps and searching for information, and establish the topological structure of the urban transportation network based on the current status.
[0070] According to the distribution of population density, transportation facilities, land use and administrative divisions in the city, the urban area to be studied is divided into an appropriate number of traffic zones (generally speaking, the number of zones in a medium-sized city does not exceed 50, and the number of zones in a large city does not exceed 100-150 at most). The occurrence and attraction of traffic trips are essentially the traffic occurrence and attraction volume of each traffic zone. When a traffic zone is the source of traffic trips, it is called the O point or starting point of the traffic trip, and when it is the attraction source of traffic trips, it is called the D point or end point of the traffic trip. Each traffic trip occurs between a pair of starting and ending points, so the starting and ending points of the trip are also called OD pairs, represented by w. W = {w} represents the set of all OD pairs in the traffic network.
[0071] Each traffic zone is connected by multiple urban roads. In the topological structure of the urban traffic network, these connecting roads are called road sections, represented by a. A = {a} represents the set of all road sections in the traffic network. By consulting maps and visiting design or construction units, important parameters such as road alignment, length and capacity can be obtained.
[0072] There are usually multiple routes to choose from between OD and w, and each route consists of multiple road segments connected end to end. In the transportation network, these routes are called paths, represented by p. w = {p} represents the set of all paths between OD pairs w. P = ∪ w P w Represents the set of all paths in the transportation network.
[0073] In addition, each travel mode in the traffic network is represented by a corresponding number i. I = {i} represents the set of all travel modes in the traffic network. Travelers in the network are divided into car users and car-free users according to whether they own a car. C represents the set of travel modes that car users can choose, and NC represents the set of travel modes that car-free users can choose. There is a relationship between the sets I = C ∪ NC.
[0074] After determining, the topology of the example network to be studied is as follows Figure 2 As shown. There are 4 traffic areas in this traffic network, marked as 1 to 4; 5 road sections, all of which are one-way, marked as ① to ⑤. Taking the OD pair (1, 2) as an example, there are three paths in this traffic network: path 1. ①→④; path 2. ②→③; path 3. ①→③→⑤.
[0075] 2) Obtain the mathematical relationship between residents’ travel demand and travel service level in the target city or region by statistically analyzing historical data or reviewing relevant cases.
[0076]
[0077]
[0078] and They represent the travel service levels felt by car users and non-car users in w, and The larger the value, the higher the travel service level. In the formula, i∈C and i∈NC represent car users and non-car users respectively; represents the generalized travel cost borne by the traveler who chooses travel mode i on path p connecting OD pair w; θ represents the parameter representing the traveler's perception error in the logit choice model.
[0079]
[0080]
[0081] and They represent the travel demands of residents with cars and without cars for OD between w. and represents the elastic demand function, which is about the travel service level and A monotonically increasing function.
[0082]
[0083]
[0084] By consulting the literature, the elastic demand function of the target city's transportation network is determined to be an exponential function. In this elastic demand function, and They represent the upper limits of OD's travel demand for car users and car-free users between w; μ represents the parameter in the exponential function, which can be obtained from traffic surveys or data fitting. Figure 3 , there is only one OD pair (1, 2) with travel demand in the target city transportation network. The upper limit of the travel demand of car users in this OD pair is The upper limit of the travel demand of car-free users is The parameter in the elastic demand function of the target city's transportation network is μ=0.5.
[0085] 3) Based on the user type and mode information, determine the travel modes existing in the target city’s transportation network.
[0086] Taking the target city's residents' travel characteristics as a reference, the corresponding travel modes in the urban transportation network include online car-hailing, private car travel and public transportation. After dividing the travel modes by roles (i.e. private car drivers, online car-hailing drivers, online car-hailing passengers, and bus passengers), the urban transportation network contains four main travel modes: private car drivers, online car-hailing drivers, online car-hailing passengers and bus passengers, corresponding to i∈SD, i∈RD, i∈R and i∈PT respectively. According to the type of travelers (users with cars, users without cars), some of the main travel modes are further subdivided to obtain RD=RD c ∪RD nc , R=R c ∪R nc and PT = PT c ∪PT nc . Where i∈R c and i∈R nc denotes online ride-hailing passengers with cars and online ride-hailing passengers without cars respectively; accordingly, i∈RD c Indicates that the online car-hailing passenger i∈R is carried c of online car-hailing drivers, i∈RD nc Indicates that the online car-hailing passenger i∈R is carried nc of online ride-hailing drivers; i∈PT c and i∈PT nc They represent bus passengers with cars and bus passengers without cars respectively.
[0087] The set of travel modes that car users in the target city’s transportation network can choose is C = SD∪RD∪R c ∪PT c , the set of travel modes that car-free users can choose is NC = R nc ∪PT nc The specific travel modes are shown in Table 1:
[0088] Table 1
[0089]
[0090] 4) Obtain the generalized travel cost parameters of the urban transportation network.
[0091] Through questionnaire surveys, data analysis and other methods, we can obtain parameters related to travelers' subjective feelings in the transportation network, such as travelers' time perception parameters and inconvenience experience parameters. By interviewing car owners, we can obtain fixed cost parameters and operating cost parameters that car users need to consider. Through online browsing and inquiries or offline visits and surveys, we can obtain parameters related to the operating mechanisms of bus companies and online car-hailing platforms, specifically: bus mileage parameters set by bus companies, online car-hailing travel time parameters, mileage parameters and floating pricing parameters set by online car-hailing platforms.
[0092] The obtained data are summarized to obtain Table 2:
[0093] Table 2
[0094]
[0095]
[0096] 5) According to the network topology, user type and mode information, a multi-mode multi-user random carpooling user equilibrium transportation network corresponding to the target city network is constructed, and carpooling matching constraints are established based on the network. The specific process is as follows:
[0097] Step 1: In the planning software or program, input the network topology obtained in step 1), label each OD pair w, each road section a, and each path p; input the elastic demand function obtained in step 2) and And the parameters involved in the elastic demand function and μ; input the numbers i of all travel modes obtained in step 3).
[0098] In the target city transportation network, x a,i represents the traffic flow of mode i on road segment a, represents the traffic flow of mode i on path p connecting OD pair w. The traffic flow of path and road segment has the following relationship:
[0099]
[0100]
[0101] Represents the segment-path conversion coefficient. When path p contains segment a, on the contrary
[0102]
[0103] The supply and demand of online ride-hailing services are based on OD pairs. represents the flow of online car-hailing drivers i∈RD between OD and w, that is, the supply of online car-hailing travel services; Denotes OD for w online ride-hailing passengers Γ r (i)∈R, that is, the demand for online car-hailing services. r (i) represents the online car-hailing passenger i∈R corresponding to the online car-hailing driver i∈RD.
[0104] Step 2: Establish ride-sharing matching constraints based on the target city’s traffic network model:
[0105]
[0106]
[0107] N i is an integer, representing the number of passengers carried in the car of online car-hailing driver i∈RD. The numerical value of i represents the specific travel mode in the target city’s transportation network. The specific situation can be found in Table 1.
[0108] 6) Based on the actual characteristics of each travel mode in the target city's transportation network, combined with the individual differences in travelers' perception of travel costs, and a series of important parameters obtained from the previous survey, a generalized travel cost suitable for the current urban sharing economy environment is constructed. The specific process is as follows:
[0109] Step 1: Obtain vehicle travel data for multiple days on each road section in the target city’s road network through traffic surveys, and deduce the travel time calculation function for the road section. When valid data cannot be obtained or the data has serious defects, the travel time in the target city’s road network can also be calculated using general empirical formulas such as the BPR function. The obtained road section travel time calculation functions are statistically summarized in Table 3.
[0110] Table 3
[0111]
[0112]
[0113] Step 2: Input the generalized travel cost parameters obtained in step 4), and combine the individual differences in travelers' perception of travel costs to conduct a refined classification of generalized travel costs from the perspectives of time perception, inconvenience experience, and actual expenses. It represents the generalized travel cost of the traveler who chooses mode i on path p of OD pair w, which is calculated as follows:
[0114]
[0115] Generalized travel costs Based on the path in the network. The symbols in the calculation formula are explained as follows:
[0116] ① Travel time perception
[0117]
[0118] Represents the travel time perceived by travelers. and They refer to the travel time of cars on path p and road section a respectively. and They refer to the bus travel time on path p and section a respectively.
[0119] ②Inconvenience experience in the car
[0120]
[0121] Indicates the degree of inconvenience experienced by online ride-hailing drivers and passengers during ride-sharing.
[0122] ③Actual travel expenses
[0123]
[0124] It indicates the cost of taking buses and online ride-hailing cars. This cost is determined by the operating mechanism of bus companies and online ride-hailing platforms, and can be queried and saved on the corresponding client. represents the mileage of path p between OD pair w. represents the Lagrange multiplier generated by the co-multiplication matching constraint. Γ rd (i) represents the online car-hailing driver i′∈RD matched by the online car-hailing passenger i∈R.
[0125] In addition, car owners will have to bear some additional costs. f represents the fixed cost parameter of owning a private car. t Represents the operating cost parameter of using a private car.
[0126] 7) Estimate the travel service level of the target city’s transportation network under the equilibrium state of multi-user random carpooling.
[0127] Step 1: In order to reflect the randomness and differences in the perception of generalized travel costs by travelers in the network, a logit-based travel choice model is used in the flow allocation process. Let the parameter θ in the logit discrete choice model = 1, and the path flow in the target city transportation network follows:
[0128]
[0129]
[0130]
[0131] Step 2: Solve the following variational inequality to obtain the path flow in the network:
[0132] In the feasible set Ω, find the vector f * ∈Ω satisfies the formula:
[0133]
[0134] In the above formula, represents the path flow vector in the network; Ω represents the feasible set of f, that is, the set of path flow vectors that satisfy the carpooling matching constraints. The specific expression is:
[0135]
[0136] in, The expression is:
[0137]
[0138] and Denote the demand function and The inverse function of is expressed in the traffic network of the target city as:
[0139]
[0140]
[0141] After substituting the parameter values, the specific calculation expression of the inverse function of the elastic demand function is:
[0142]
[0143]
[0144] f * represents the solution of the variational inequality, that is, the traffic flow of the city's transportation network under the equilibrium state of multi-user random carpooling. The solution process can use the PATH solver of GAMS software, or a suitable convergence algorithm. The flow distribution in the transportation network of the target city is shown in Table 4. According to the travel service level calculation formula, the travel service level of the target city's transportation network can be obtained.
[0145] Table 4
[0146]
[0147]
[0148]
[0149]
[0150] 8) Based on the elastic demand function and the obtained travel service level, predict the travel demand of residents in the complex transportation network of the target city. The specific calculation is:
[0151]
[0152]
[0153] It can be concluded that in the transportation network of the target city, the travel demand of residents with cars is 1.010, and the travel demand of residents without cars is 1.803.
[0154] The embodiments are only for illustrating the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.
Claims
1. A method for predicting residents' travel demand for urban shared transportation, characterized in that: The following steps are involved: (1) Establish the topological structure of the urban transportation network. The travel modes in the urban transportation network include private car travel, online car-hailing travel, and public transportation travel. The user types in the urban transportation network include car users and non-car users. Obtain the perception parameters of travelers and the operating parameters of public transportation companies and online car-hailing platforms; (2) Obtain the functional relationship between urban residents’ travel demand and travel service level based on historical data, namely, the elastic demand function; (3) Based on the topological structure of the urban transportation network, user types, and travel mode information, a multi-mode multi-user random carpooling user equilibrium transportation network corresponding to the urban network is constructed, and carpooling matching constraints are established based on the network; (4) Based on the actual characteristics of each travel mode in the urban transportation network and the individual differences in travelers' perception of travel costs, a generalized travel cost applicable to the shared travel environment is constructed; (5) Estimate the travel service level of the city's transportation network under the equilibrium state of multi-user random carpooling; (6) Based on the elastic demand function, combined with the obtained travel service level of the transportation network and the generalized travel cost under this service level, predict the travel demand of residents in the urban transportation network; In step (2), the functional relationship between the urban residents' travel demand and the travel service level is as follows: In the above formula, and They represent the travel demands of residents with and without cars between w by OD; and represents the elastic demand function; and They represent the upper limits of OD's travel demand for car users and car-free users between w; μ represents the parameter in the exponential function; and They represent the travel service levels felt by car users and non-car users in w, and The larger the value, the higher the level of travel service: In the above formula, i∈C and i∈NC represent car users and car-free users respectively; represents the generalized travel cost borne by the traveler who chooses travel mode i on the path p connecting the OD pair w; θ represents the parameter representing the traveler's perception error in the logit choice model; The specific process of step (3) is as follows: (301) Inputting the topological structure, user type and travel mode information of the urban road network, and constructing a multi-mode multi-user urban transportation network structure corresponding to the urban transportation network in the software, the transportation network structure including the connection status of nodes and road sections in the road network, and the road section capacity and length information; In terms of user type, according to whether the travelers in the network own private cars, they are divided into car users and car-free users, represented by i∈C and i∈NC respectively; In terms of travel modes, the three modes of online car-hailing, private car travel and bus travel in the transportation network are divided according to travel roles, resulting in four travel modes: private car drivers, online car-hailing drivers, online car-hailing passengers and bus passengers, corresponding to i∈SD, i∈RD, i∈R, i∈PT respectively; the above travel modes are further subdivided according to whether or not a person owns a private car: RD=RD c ∪RD nc , R=R c ∪R nc and PT = PT c ∪PT nc ; where i∈R c and i∈R nc They represent online ride-hailing passengers with cars and online ride-hailing passengers without cars respectively; i∈RD c Indicates that the online car-hailing passenger i∈R is carried c of online car-hailing drivers, i∈RD nc Indicates that the online car-hailing passenger i∈R is carried nc of online ride-hailing drivers; i∈PT c and i∈PT nc They represent bus passengers with cars and bus passengers without cars respectively; (302) Based on the constructed multi-mode and multi-user urban transportation network structure, the ride-sharing matching constraints are established: In the above formula, represents the flow of travelers who choose mode i on path p connecting OD pair w, Γ r (i) represents the online car-hailing driver i∈RD corresponding to the online car-hailing passenger i∈R, N i represents the number of passengers in the car of online car-hailing driver i; In the above formula, represents the flow of online car-hailing drivers i∈RD between OD and w, that is, the supply of online car-hailing travel services; Denotes OD for w online ride-hailing passengers Γ r (i) ∈R, i.e., the demand for online ride-hailing services; In step (4), the generalized travel cost is: In the above formula, Represents the travel time perceived by travelers: In the above formula, ρ i refers to the time value parameter of the traveler who chooses mode i, and They refer to the driving time of cars on path p and road section a respectively. and They refer to the bus travel time on path p and section a respectively, Represents the segment-path conversion coefficient. If path p contains segment a, otherwise, Expressing inconvenience experience in the car: In the above formula, γ i is the inconvenience experience parameter of travelers who choose mode i; Express the cost of travellers: In the above formula, represents the mileage of path p between OD pairs w, b i , and r i They are the travel time parameter, mileage parameter and floating pricing parameter set by the online car-hailing platform, τ b It represents the bus mileage parameters set by the bus company. represents the Lagrange multiplier generated by the co-multiplication matching constraint, Γ rd (i) represents matching the online car-hailing passenger i∈R to the responding online car-hailing driver; c f represents the fixed cost of a private car, c t Represents the cost of using a private car.
2. The method for predicting residents' travel demand for urban shared transportation according to claim 1 is characterized in that: The specific process of step (5) is as follows: (501) Considering the randomness and differences in travelers’ perception of generalized travel costs, a logit-based travel choice model is used in the flow allocation process, and the path flow is obtained as follows: In the above formula, represents the probability that a traveler between OD pair w chooses mode i on path p; (502) In the feasible set Ω, find the vector f * ∈Ω satisfies: In the above formula, represents the path flow vector in the network, W is the set of all OD pairs in the urban transportation network, P w is the set of all paths between OD pairs w, I represents the set of all travel modes; Ω represents the set of path flow vectors that satisfy the ride-sharing matching constraints; vector The specific expression is: in, The expression is: and Denote the demand function and The inverse function of is expressed as: f * represents the solution of the variational inequality, that is, the traffic flow of the urban transportation network under the equilibrium state of multi-user random carpooling; based on this, the travel service level of the urban transportation network under the equilibrium state of multi-user random carpooling is obtained. and 3. The method for predicting residents' travel demand for urban shared transportation according to claim 2 is characterized in that: In step (6), according to the travel service level obtained in step (5) and Combined with the elastic demand function obtained in step (2) and Predict the travel needs of residents in the city's transportation network:
Citation Information
Patent Citations
Urban multi-mode traffic network transportation capacity coordination evaluation method
CN110414833A
Combined prediction method for short-time travel requirements of online hailed car
CN110415013A
