Railway traffic distribution method based on cellular network gradient algorithm

Abstract

The invention discloses a railway traffic distribution method based on a cellular network gradient algorithm. The method comprises the following steps: (A) vectorizing a railway basic network; (B) traffic cell construction based on a cellular network; (C) extracting a trip chain based on cellular network big data; (D) establishing a railway traffic flow distribution model based on a cellular network gradient algorithm; and (E) solving the model. According to the method, the mathematical model established through the cellular network gradient algorithm can accurately describe the topological structure and the operation rule of the railway network, constraint conditions such as train operation time, speed and capacity are considered, and the target function for maximizing the throughput and the operation efficiency of the railway network is also considered, so that the distribution of railway traffic flow can be effectively optimized.

Description

Technical Field

[0001] The present invention belongs to the field of integrated transportation planning and management, and in particular relates to a railway traffic allocation method based on a cellular network gradient algorithm. Background Art

[0002] Traffic volume forecasting is the main basis for railway system planning, project investment and construction, and traffic distribution is the core of traffic volume forecasting. In the context of multi-network integration, passenger flow is more sensitive to punctuality, comfort and other perceptions, and passenger flow differences are enhanced. The traditional passenger flow distribution uses an aggregate model that is difficult to reflect individual differences, and its punctuality, comfort and other aspects are difficult to quantify accurately, and the impedance calculation accuracy is insufficient. With the development of mobile communication big data technology, mobile phone users will record the base station numbers of their services that change over time in the database of mobile operators during their travels. The present invention targets the spatial characteristics of railways, utilizes the cellular communication network topology structure, a network construction concept in mobile communication networks, extracts railway travel chains based on cellular network big data, and establishes a railway traffic distribution model based on the cellular network gradient algorithm to achieve rapid distribution of traffic flow on the railway network. Summary of the Invention

[0003] Purpose of the invention: To address the deficiencies of the existing technology, a railway traffic allocation method based on a cellular network gradient algorithm is provided. This method utilizes massive amounts of mobile phone signaling data to extract residents' railway travel chains, and then establishes a railway traffic allocation model based on a cellular network gradient algorithm. This achieves rapid allocation of traffic flow on the railway network, greatly improving the accuracy and speed of allocation, and providing strong support for railway project planning and construction.

[0004] Technical solution: In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A railway traffic assignment method based on a cellular network gradient algorithm comprises the following steps:

[0006] (A) Railway basic network vectorization;

[0007] (B) Construction of traffic cells based on cellular networks;

[0008] (C) Trip chain extraction based on cellular network big data;

[0009] (D) Establish a railway traffic flow distribution model based on cellular network gradient algorithm;

[0010] (E) Model solution.

[0011] Preferably, the railway infrastructure network vectorization in step (A) comprises the following steps:

[0012] (A1) Based on the geographical location information and station attribute information of the railway station, the railway station is vectorized and a railway station information table is constructed. The data structure is shown in the following table:

[0013] Table 1 Schematic diagram of station information data storage structure

[0014] Serial number Station Name longitude latitude 1 Xi'an Station 23.0024 113.3275 ..... ..... ..... .....

[0015] (A2) Based on the geographical location information and line attribute information of the railway line, the railway line is appropriately quantified and a railway line information table is constructed, which mainly includes a line basic information table and a line topology point coordinate table, as shown below:

[0016] Table 2 Schematic diagram of data storage structure for basic line information

[0017]

[0018] Table 3 Schematic diagram of data storage structure of line topology point information

[0019]

[0020]

[0021] Preferably, the construction of the traffic cell based on the cellular network in step (B) is based on the cell group calculation principle, with reference to the cell group construction method and standard in the communication field, and takes 1 / 2 of the shortest distance between rail transit stations as the cell radius to construct the traffic cell based on the cellular network big data, specifically as follows: Figure 2 shown.

[0022] Preferably, the travel chain extraction based on cellular network big data in step (C) is to match the spatiotemporal trajectory of residents' travel with the railway network constructed in step (A) based on the mobile phone signaling data of residents' travel, that is, to convert the spatiotemporal trajectory sequence of residents (data structure shown in Table 4) into a railway section sequence (data structure shown in Table 5);

[0023] Table 4 Schematic diagram of the data storage structure of residents' travel time and space trajectory

[0024] time longitude latitude 8：00 110.8996931 130.6110508 8：05 130.63968 127.6635108 8：10 130.63968 127.6635108 ..... ..... .....

[0025] Table 5 Schematic diagram of data storage structure for railway sections of residents' travel

[0026] time Railway section number 8：00 a1 8：05 a2 8：10 a2 ..... .....

[0027] Preferably, the railway traffic flow distribution model based on the cellular network gradient algorithm described in step (D) can be expressed by the following formula:

[0028]

[0029] Where:

[0030] Z(x): total system impedance

[0031] q rs : OD flow between the starting point r and the end point s

[0032] OD traffic on the kth path of rs

[0033] x a : Traffic flow on section a

[0034] t a (x a ): Impedance function of section a with flow rate as the independent variable

[0035] OD is the impedance of the kth path to RS

[0036] If segment a is on path k connecting OD pair rs, the value is 1; otherwise, it is 0.

[0037] D a : The passing capacity of section a

[0038] Preferably, the impedance function t a (x a ) can be calculated using the following formula:

[0039]

[0040] Where:

[0041] d(a): length of road section a

[0042] v(a): average travel speed on road section a

[0043] f h : Passenger unit time value

[0044] f a : Freight rate per unit distance

[0045] w h ,w f : Weight coefficient of travel time and travel cost, w h +w f =1

[0046] t0: travel time of section a

[0047] X a : Maximum capacity limit of section a

[0048] μ, β: constants, obtained through regression, with reference values of 2.6 and 1 respectively

[0049] Preferably, the model solution in step (E) may adopt a gradient descent algorithm to obtain the optimal traffic flow allocation solution. The specific steps are as follows:

[0050] (1) Initialize the impedance of section a Perform all-or-nothing (0-1) traffic distribution to obtain the traffic flow of each section x

[0051] (2) Update the impedance of each section:

[0052] (3) Search for the gradient vector direction of the objective function, the gradient function of the total impedance of the system: According to the updated impedance of each section, the 0-1 flow distribution is redistributed. If there is an existing section L1, the exploration direction is obtained with the additional flow (a group of sections with additional flow).

[0053] (4) Search for the gradient vector direction of the objective function

[0054] when (D is the traffic capacity of the road section), repeat steps (2) and (3).

[0055] when (D is the capacity of the road section), search the gradient vector direction of the objective function, and follow the updated impedance of each road section. If there is a planned road section L2, perform 0-1 flow distribution to obtain the additional flow in the exploration direction Proceed to the next step.

[0056] (5) One-dimensional search, the formula Substitute into the objective function and use the bisection method to find the exploration step size.

[0057] (6) Convergence judgment: but This is the solution, and the calculation ends. Otherwise, n=n+1, go to step (2). BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is the overall flow chart of the present invention;

[0059] Figure 2 This is a schematic diagram of traffic cell division based on cellular network;

[0060] Figure 3 Schematic diagram of traffic zone division based on cellular network in Xi'an. DETAILED DESCRIPTION

[0061] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0062] (A) Railway basic network vectorization;

[0063] (A1) Based on the geographical location information and station attribute information of the railway station, the railway station is vectorized and a railway station information table is constructed:

[0064] Table 1 Station information table

[0065] Serial number Station Name longitude latitude 1 Beidajie Station 106.1032561 114.8925021 2 Bell Tower Station 107.2399195 121.8935095 3 Nanshaomen Station 147.2798244 127.0614293 ..... ..... ..... .....

[0066] (A2) Based on the geographical location information and line attribute information of the railway line, the railway line is appropriately quantified and a railway line information table is constructed, which mainly includes a line basic information table and a line topology point coordinate table, as shown below:

[0067] Table 2 Schematic diagram of data storage structure for basic line information

[0068]

[0069] Table 3 Schematic diagram of data storage structure of line topology point information

[0070] Topological point number longitude latitude 4 41.71433545 26.6851758 5 18.66157093 6.773907146 ..... ..... .....

[0071] (B) Construction of traffic cells based on cellular networks;

[0072] The shortest distance between rail transit stations is half of the cell radius, and Xi'an's traffic community based on cellular network big data is constructed. Figure 3 shown.

[0073] (C) Trip chain extraction based on cellular network big data;

[0074] The travel trajectory map matching method is used to convert the residents' spatiotemporal trajectory sequence into a railway section sequence.

[0075] Table 4 Schematic diagram of the data storage structure of residents' travel time and space trajectory

[0076] time longitude latitude 8：00 110.8996931 130.6110508 8：05 130.63968 127.6635108 8：10 130.63968 127.6635108 ..... ..... .....

[0077] Table 5 Schematic diagram of data storage structure for railway sections of residents' travel

[0078] time Railway section number 8：00 a1 8：05 a2 8：10 a2 ..... .....

[0079] (D) Establish a railway traffic flow distribution model based on cellular network gradient algorithm;

[0080]

[0081] The impedance function Substitute into the model:

[0082]

[0083]

[0084] (E) Model solution;

[0085] (1) Initialize the impedance of section a Perform all-or-nothing (0-1) traffic distribution to obtain the traffic flow of each section

[0086] (2) Update the impedance of each section:

[0087] (3) Search for the gradient vector direction of the objective function, the gradient function of the total impedance of the system: According to the updated impedance of each section, the 0-1 flow distribution is redistributed. If there is an existing section L1, the exploration direction is obtained with the additional flow (a group of sections with additional flow).

[0088] (4) Search for the gradient vector direction of the objective function

[0089] when (D is the traffic capacity of the road section), repeat steps (2) and (3).

[0090] when (D is the capacity of the road section), search the gradient vector direction of the objective function, and follow the updated impedance of each road section. If there is a planned road section L2, perform 0-1 flow distribution to obtain the additional flow in the exploration direction Proceed to the next step.

[0091] (5) One-dimensional search, the formula Substitute into the objective function and use the bisection method to find the exploration step size.

[0092] (6) Convergence judgment: but This is the solution, and the calculation ends. Otherwise, n=n+1, go to step (2).

[0093] After the allocation is completed, the traffic flow of each railway section is obtained, as shown in the following table:

[0094] Table 6 Railway section flow data after allocation

[0095]

[0096] The above are only preferred embodiments of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A railway traffic assignment method based on a cellular network gradient algorithm, characterized in that: The following steps are involved: (A) Railway basic network vectorization; (B) Construction of traffic cells based on cellular networks; (C) Trip chain extraction based on cellular network big data; (D) Establish a railway traffic flow distribution model based on cellular network gradient algorithm; (E) Model solution.

2. The railway traffic assignment method based on cellular network gradient algorithm according to claim 1, characterized in that: The railway basic network vectorization described in step (A) includes the following steps: (A1) vectorizing the railway station based on the geographical location information and station attribute information of the railway station and constructing a railway station information table; (A2) According to the geographical location information and line attribute information of the railway line, the railway line is appropriately quantified and a railway line information table is constructed.

3. The railway traffic assignment method based on cellular network gradient algorithm according to claim 1, characterized in that: The construction of the traffic cell based on the cellular network described in step (B) is based on the cell cluster calculation principle, refers to the cell cluster construction method and standards in the communication field, and takes 1 / 2 of the shortest distance between rail transit stations as the cell radius to construct a traffic cell based on cellular network big data.

4. The railway traffic assignment method based on cellular network gradient algorithm according to claim 1, characterized in that: The travel chain extraction based on cellular network big data described in step (C) is to match the spatiotemporal trajectory of residents' travel with the railway network constructed in step (A) based on the mobile phone signaling data of residents' travel, that is, to convert the spatiotemporal trajectory sequence of residents into a railway section sequence.

5. The railway traffic assignment method based on cellular network gradient algorithm according to claim 1, characterized in that: The railway traffic flow distribution model based on the cellular network gradient algorithm described in step (D) can be expressed as follows: Where: Z(x): total system impedance q rs : OD flow between the starting point r and the end point s OD traffic on the kth path of rs x a : Traffic flow on section a t a (x a ): Impedance function of section a with flow rate as the independent variable OD is the impedance of the kth path to RS If segment a is on path k connecting OD pair rs, the value is 1; otherwise, it is 0. D a : The carrying capacity of section a.

6. The railway traffic flow distribution model based on cellular network gradient algorithm according to claim 5 is characterized in that: The impedance function t a (x a ) can be calculated using the following formula: Where: d(a): length of road section a v(a): average travel speed on road section a f h : Passenger unit time value f a : Freight rate per unit distance w h ,w f : Weight coefficient of travel time and travel cost, w h +w f =1 t0: travel time of section a X a : Maximum capacity limit of section a μ, β: constants.

7. The railway traffic assignment method based on cellular network gradient algorithm according to claim 1, characterized in that: The model solution described in step (E) can use a gradient descent algorithm.

Images