CN113077079A - Data-driven rail transit new line access passenger flow prediction method - Google Patents
Data-driven rail transit new line access passenger flow prediction method Download PDFInfo
- Publication number
- CN113077079A CN113077079A CN202110312209.1A CN202110312209A CN113077079A CN 113077079 A CN113077079 A CN 113077079A CN 202110312209 A CN202110312209 A CN 202110312209A CN 113077079 A CN113077079 A CN 113077079A
- Authority
- CN
- China
- Prior art keywords
- station
- passenger flow
- data
- index
- rail transit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000013139 quantization Methods 0.000 claims abstract description 35
- 238000007619 statistical method Methods 0.000 claims abstract description 13
- 238000000611 regression analysis Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 7
- 230000010354 integration Effects 0.000 claims description 38
- 238000012546 transfer Methods 0.000 claims description 20
- 230000008859 change Effects 0.000 claims description 17
- 230000001419 dependent effect Effects 0.000 claims description 15
- 230000000694 effects Effects 0.000 claims description 13
- 230000000877 morphologic effect Effects 0.000 claims description 12
- 238000010219 correlation analysis Methods 0.000 claims description 10
- 230000011218 segmentation Effects 0.000 claims description 9
- 238000013461 design Methods 0.000 claims description 8
- 238000010586 diagram Methods 0.000 claims description 8
- 238000004140 cleaning Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000013501 data transformation Methods 0.000 claims description 6
- 238000012417 linear regression Methods 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 3
- 238000011835 investigation Methods 0.000 abstract description 6
- 238000010276 construction Methods 0.000 description 5
- 230000006399 behavior Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The application relates to a data-driven rail transit new line access passenger flow prediction method. The method comprises the following steps: acquiring characteristic data of a rail transit station and passenger flow data of the rail transit station of a target rail station; carrying out quantization processing on the characteristic data of the rail transit station to obtain station characteristic quantization indexes of a target rail station; carrying out statistical analysis on the passenger flow data of the rail transit station to obtain a station passenger flow characteristic index of a target rail station; taking a site characteristic quantization index as an autovariable set, taking a site passenger flow characteristic index as a cause variable set, and constructing a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model; carrying out parameter calibration on the model by adopting a regression analysis method, and determining the parameter calibrated by the prediction model; the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model predict according to the parameters calibrated by the prediction model to obtain the passenger flow total amount and the passenger flow fluctuation form of the target track station, and the problems of high investigation cost and low prediction precision are solved.
Description
Technical Field
The application relates to the technical field of road traffic, in particular to a data-driven method for predicting new line access passenger flow of rail transit.
Background
Urban rail transit construction cost is high, and the construction cycle is long, and a city is difficult to once only build large-scale rail transit net, consequently each city develops the process that rail transit all is along with the road network and enlarges gradually. When a new line (the new line refers to a newly opened rail transit line) is connected into a line network and is put into operation, the topology structure of the original urban rail transit line network is changed by the new line connection, and meanwhile, the increase of the site density enables travelers to have various choices of adjacent sites, so that the traveling selection behavior of the travelers is influenced, and the distribution of passenger flow on the line network is changed. The method is necessary for better mastering the passenger flow rule of the line network, formulating a reasonable rail service operation plan, improving the operation efficiency of the line network and providing data support for formulating rail traffic demand management measures and predicting the passenger flow after the new line is accessed.
The new line can not analyze and predict the rule through historical synchronization data due to the lack of historical data of the new line. At the present stage, a four-stage prediction model method based on travel is generally adopted for predicting passenger flow accessed by a new line, and the method analyzes the current urban situation and the future traffic situation according to four stages of traffic generation prediction, traffic distribution prediction, traffic mode division prediction and traffic distribution, and is the method which is most widely applied in the current traffic planning field. However, this method has problems of high investigation cost and low prediction accuracy.
Disclosure of Invention
In view of the above, it is necessary to provide a method for predicting new line access passenger flow of rail transit based on data driving, which can solve the problems of high investigation cost and low prediction accuracy.
A data-driven rail transit new line access passenger flow prediction method comprises the following steps:
acquiring rail transit station characteristic data and rail transit station passenger flow data of a target rail station;
quantizing the rail transit station characteristic data to obtain a station characteristic quantization index of the target rail transit station;
carrying out statistical analysis on the passenger flow data of the rail transit station to obtain a station passenger flow characteristic index of the target rail station;
taking the station characteristic quantitative index as a self-variable set, and taking the station passenger flow characteristic index as a cause variable set to construct a passenger flow total prediction model and a passenger flow fluctuation form prediction model;
carrying out parameter calibration on the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by adopting a regression analysis method, and determining a parameter calibrated by the prediction model;
and the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model predict according to the parameters calibrated by the prediction model to obtain the passenger flow total amount and the passenger flow fluctuation form of the target track station.
In one embodiment, the step of performing quantization processing on the rail transit station characteristic data to obtain a station characteristic quantization index of the target rail transit station includes:
quantifying according to the land utilization related data in the rail transit station characteristic data to obtain an attraction range activity index of the target rail transit station;
quantifying according to station continuing design related data in the rail transit station characteristic data to obtain a bus transfer index, a shared single bus transfer index and a motor vehicle parking point index of the target rail transit station;
quantifying according to station reachability relevant data in the rail transit station characteristic data to obtain an inter-station reachability index and a station domain reachability index of the target rail station;
the attraction range activity index, the bus transfer index, the shared bicycle transfer index, the motor vehicle parking point index, the inter-station reachability index and the station domain reachability index form the station characteristic quantization index.
In one embodiment, the formula for quantifying according to the land use related data in the rail transit station characteristic data is as follows:
wherein, the influence range of the target track site is totally k POIs, PjIs the attraction range activity of the target track station j, NmjIs the number of m-th POIs within the influence range of the target track site j, AmjIs the rate of travel, Square, of POIs of class m within the influence of target track site jjIs the area of influence of the target track station j.
In one embodiment, the step of obtaining an inter-station reachability index and a station domain reachability index by performing quantization according to station reachability related data in the rail transit station characteristic data includes:
acquiring a track traffic line network diagram and a street network from the station reachability related data in the track traffic station characteristic data;
converting the track traffic line network diagram into a station topological graph taking the target track station as a core by performing convex space segmentation on the track traffic line network diagram;
performing spatial morphological analysis based on the station topological graph to obtain the integration level and the selectivity of the target track station;
carrying out weighted average on the integration degree and the selectivity degree to obtain an inter-station reachability index of the target track station;
abstracting the street network into a space system formed by axes based on an axis segmentation method;
calculating a global integration average value of streets directly connected with the target track station and a local integration average value of 600-meter radius of all streets in the target track station domain based on an axis relation of a space system to obtain the global integration average value and the local integration average value;
and carrying out weighted average on the global integration average value and the local integration average value to obtain the station domain reachability index of the target track station.
In one embodiment, the step of performing statistical analysis on the rail transit station passenger flow data to obtain a station passenger flow characteristic index of the target rail transit station includes:
carrying out data cleaning and data transformation on the rail transit station passenger flow data to obtain rail transit station time sequence data;
performing statistical analysis according to the track station time sequence data to obtain structural feature data and morphological feature data of the target track station;
and the structural characteristic data and the morphological characteristic data are used as station passenger flow characteristic indexes.
In one embodiment, the step of constructing a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model by using the site characteristic quantization index as an autovariant set and using the site passenger flow characteristic index as a cause variant set includes:
analyzing the correlation between the dependent variable set and the independent variable set by taking the station characteristic quantization index as an independent variable set and taking the station passenger flow characteristic index as a dependent variable set to obtain a correlation analysis result;
according to the correlation analysis result, a passenger flow basic prediction model, a passenger flow total amount trend change model and a passenger flow fluctuation form prediction model are constructed;
and fusing the passenger flow basic prediction model and the passenger flow total amount trend change model to obtain a passenger flow total amount prediction model.
In one embodiment, the basic passenger flow prediction model is:
wherein h isjQj(0)And (3) obtaining a passenger flow basic total amount prediction result of the target track station j, wherein epsilon is a regression intercept, A is a regression coefficient, and IV is an autovariate set.
In one embodiment, the passenger flow volume trend change model is as follows:
wherein the content of the first and second substances,
t represents the difference between the year of new line operation and the year of prediction of new line access passenger flow,
and predicting the result of the total passenger flow of the target track station j by considering the growth rate.
In one embodiment, the total passenger flow prediction model is:
wherein Q isjAnd predicting the result of the total passenger flow of the target track station.
In one embodiment, the passenger flow fluctuation form prediction model is as follows:
DVj＝AqpIVp+ε
wherein, DVjIs the result of the prediction of the traffic fluctuation pattern of the target track station j, ApqIndependent variable IV with number p in linear regression equation for dependent variable with number qpCoefficient of (IV)pIs an argument with sequence number p.
According to the data-driven rail transit new line access passenger flow prediction method, rail transit station characteristic data and rail transit station passenger flow data of a target rail transit station are obtained; carrying out quantization processing on the characteristic data of the rail transit station to obtain station characteristic quantization indexes of a target rail station; carrying out statistical analysis on the passenger flow data of the rail transit station to obtain a station passenger flow characteristic index of a target rail station; taking a site characteristic quantization index as an autovariable set, taking a site passenger flow characteristic index as a cause variable set, and constructing a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model; carrying out parameter calibration on the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by adopting a regression analysis method, and determining the calibrated parameters of the prediction model; the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model are used for predicting according to the parameters calibrated by the prediction model to obtain the passenger flow total amount and the passenger flow fluctuation form of the target railway station, complex data investigation is not needed, the operability is higher, the station passenger flow total amount and the passenger flow fluctuation form are predicted according to the influence factors of the railway traffic passenger flow rule based on the historical data of the existing station, the theoretical basis is higher, the precision is higher, and the problems of high investigation cost and low prediction precision are solved.
Drawings
Fig. 1 is a schematic flow chart of a method for predicting new line access passenger flow of rail transit based on data driving in one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In one embodiment, as shown in fig. 1, there is provided a data-driven rail transit new line access passenger flow prediction method, including the following steps:
step S220, acquiring the rail transit station characteristic data and the rail transit station passenger flow data of the target rail station.
Wherein the target track station is a predicted track traffic station. The rail transit station characteristic data comprises land utilization related data, station connection design related data, station accessibility related data and the like. The land utilization related data comprises data such as the number of various POIs in the influence range of the rail transit station; the station connection design related data comprises data such as the number of bus stations in the influence range of the rail transit stations, the number of shared single vehicle parking points in the influence range of the rail transit stations, the number of motor vehicle parking lots in the influence range of the rail transit stations, the nearest distance between the motor vehicle parking lots in the influence range of the rail transit stations and the exit of the rail transit stations and the like; the station accessibility related data comprises data of a rail transit network diagram, a street network and the like. The rail transit station passenger flow data is rail passenger flow card swiping data of a target rail station.
And step S240, carrying out quantization processing on the rail transit station characteristic data to obtain a station characteristic quantization index of the target rail transit station.
The station characteristic quantitative indexes come from land utilization around the target track station, station connection design and station accessibility, and comprise an attraction range activity index P and a bus transfer index T of the target track stationbusShared bicycle transfer index TbiMotor vehicle parking point index TpAn inter-station reachability index, a station domain reachability index, and the like. The influence range of the target track station is selected from a superposition influence range (possibly, the same station has a plurality of exit ports) which takes each exit port of the target track station as a core and takes 500m as a radius. And determining a station characteristic quantization index based on the influence range of the target track station.
In one embodiment, the step of performing quantization processing on the rail transit station characteristic data to obtain a station characteristic quantization index of a target rail transit station includes:
quantifying according to the land utilization related data in the rail transit station characteristic data to obtain an attraction range activity index of the target rail station; quantifying according to station continuing design related data in the rail transit station characteristic data to obtain a bus transfer index, a shared single bus transfer index and a motor vehicle parking point index of a target rail station; quantifying according to station reachability relevant data in the rail transit station characteristic data to obtain an inter-station reachability index and a station domain reachability index of a target rail station; the attraction range activity index, the bus transfer index, the shared single-bus transfer index, the motor vehicle parking point index, the inter-station accessibility index and the station domain accessibility index of the target track station form a station characteristic quantization index.
Aiming at the utilization of the peripheral land of the target track site, the number of various POIs in the influence range of the target track site is selected as the attraction range activity index of the target track site. Specifically, the attraction range activity index of the target track site is determined according to the POI information in a large range around the target track site, and the data targeting various types (types, longitude and latitude) of POI data can be acquired by combining with an API open platform provided by a map navigation service provider, for example, the POI information in a large range around the target track site is collected by data crawling, and the POI data in a non-site influence range is screened and removed in the later data preprocessing.
Aiming at the continuing design of the target track station, a bus transfer index, a shared single-bus transfer index and a motor vehicle parking point index are selected to represent the influence of the continuing design of the target track station on the passenger flow of the station. The bus transfer index is determined based on the number of bus stops in the influence range of the target track stop, and is represented as Tbus. The shared single-car transfer index is determined based on the number of shared single-car parking points in the influence range of the target track station, and is represented as Tbi. The motor vehicle parking point index is determined based on the number of motor vehicle parking lots in the target track station influence range and the nearest distance between the motor vehicle parking lots in the target track station influence range and the exit, and is represented as TpSpecifically:
wherein N ispNumber of motor vehicle parking lots within the influence range of the target track station, DpThe distance between the vehicle parking lot closest to the exit port and the exit port is the distance between all the vehicle parking lots in the influence range of the target track station.
In one embodiment, the formula for quantifying according to land use related data in the rail transit station characteristic data is:
wherein, the influence range of the target track site is totally k POIs, PjIs the attraction range activity of the target track station j, NmjIs m-th type POI at targetNumber of track stations j within influence, AmjIs the trip rate of the m-th POI in the influence range of the target track site j (the trip rate refers to the construction project trip rate, and the unit is the number of people/100 m2The building area can be determined by referring to the related regulations of construction project traffic influence evaluation of the Ministry of housing and construction), SquarejIs the area of influence of the target track station j.
In one embodiment, the step of obtaining the inter-station reachability index and the station domain reachability index of the target track station by performing quantization according to the station reachability related data in the track traffic station characteristic data includes: acquiring a track traffic line network diagram and a street network from station reachability related data in the track traffic station characteristic data; the method comprises the steps that convex space segmentation is carried out on a track traffic line network graph, and the track traffic line network graph is converted into a station topological graph with a target track station as a core; performing spatial morphological analysis based on the station topological graph to obtain the integration level and the selectivity of the target track station; carrying out weighted average on the integration degree and the selectivity degree to obtain an inter-station reachability index of a target track station; abstracting a street network into a space system formed by axes based on an axis segmentation method; calculating the global integration average value of streets directly connected with the target track station and the local integration average value of all the streets in the station domain at the radius of 600 meters based on the axis relation of the space system to obtain the global integration average value and the local integration average value; and carrying out weighted average on the global integration average value and the local integration average value to be used as an index for measuring the station domain reachability of the target track station.
And aiming at the accessibility of the stations, selecting an inter-station accessibility index and a station domain accessibility index to comprehensively represent the accessibility of the target track station. The inter-station accessibility index firstly acquires a rail transit network map, each rail transit station in the rail transit network map is regarded as a convex space, line sections among the rail transit stations are abstracted into connection relations among the convex spaces, the rail transit network map is subjected to convex space segmentation, after the convex space segmentation is completed, the rail transit network map is converted into a station topological map with a target rail station as a core, the station is represented by a circle, the line section is represented by a short line, then, the spatial form analysis is carried out according to the station topological map, and spatial form variables are obtained and comprise an integration level I and a selectivity C.
Integration level: summing and negating the shortest topological steps of the target track station to other track traffic stations in the whole track network, and eliminating the scale interference and the symmetry interference of the network to obtain the integration level of the target track station, wherein the integration level of the target track station is represented as Ij。
Wherein n is the total number of railway traffic line network stations, LjlThe number of the shortest topological steps between the target track station j and the other track traffic stations l of the track traffic network is referred to.
The selection degree is as follows: on the premise of eliminating the target track station in the track traffic network, sequentially selecting the rest track traffic stations as starting points and recording the starting points as OxSequentially removing the starting point O from the rest rail transit stationsxAs a starting point OxEnd point of (D)yWhere x is 1, 2, …, n-1, y is 1, 2, …, n-1, x ≠ y, determined from the starting point OxReaches the end point DyAnd calculating the number of times that the target track station j exists in the shortest path.
In the formula, CjIs the selectivity of a target track station j, n is the total number of track traffic stations of the track traffic network, j is the target track station, L [ O ]x,Dy]Is a starting point in a rail transit network as OxEndpoint is DyOf the shortest path, PjIs an intermediate amount of selectivity for the target track site j.
And carrying out weighted average on the integration degree and the selection degree to be used as an index for measuring inter-station accessibility of the target track station.
Station domain reachability: firstly, abstracting a street network into a space system formed by the axes based on an axis segmentation method, namely: covering the entire street network with the smallest and longest axes, forming a spatial system of axes, wherein each axis corresponds to a node; second, the global integration mean of streets directly connected to the target track site is calculated. The calculation formula is as follows:
wherein, IjRallIs the global integration mean, I, of streets directly connected to the target track sitejtRallIs the global integration of the streets directly connected to the target track sites, N is the number of nodes, MDtIs the average depth value of the node t, dtuIs the shortest distance from node t to other nodes u, RAiIs a relative asymmetry value, RRAiFor practical relative asymmetry values, D is derived from the diamond-shaped topological structure model and is used to normalize the integration.
Thirdly, calculating the average value of the local integration degree of all the streets of the station domain of the target track station within 600 m of the radius. The calculation formula is as follows:
wherein, IjR600Is the local integration average value of 600 m radius of all streets in the station domain of the target track station, IjtR600Is the local integration degree of all streets with the radius of 600 meters in the station domain of the target track station, N600Is the number of nodes in the 600-meter radius of all streets in the station domain.
And finally, carrying out weighted average on the global integration average value of the streets directly connected with the target track station and the local integration average value of all 600-meter radiuses of all the streets in the station domain of the target track station, wherein the weighted average is used as an index for measuring the station domain accessibility of the target track station.
And step S260, carrying out statistical analysis on the passenger flow data of the rail transit station to obtain the station passenger flow characteristic index of the target rail station.
In one embodiment, the step of performing statistical analysis on the rail transit station passenger flow data to obtain a station passenger flow characteristic index of a target rail station includes: carrying out data cleaning and data transformation on passenger flow data of the rail transit station to obtain time sequence data of the rail transit station; performing statistical analysis according to the time series data of the track station to obtain structural characteristic data and morphological characteristic data of the target track station; and taking the structural characteristic data and the morphological characteristic data as station passenger flow characteristic indexes of the target track station.
The data for carrying out data cleaning and data transformation on the passenger flow data of the rail transit station specifically comprises the following steps: for other data records contained in the rail transit station passenger flow data, corresponding fields can be added by combining the data content of the other data records, or the other data records are directly deleted from the card swiping data, and the rail transit station passenger flow data is subjected to data cleaning by combining the characteristics of the rail transit station passenger flow card swiping data, so that the cleaned rail transit station passenger flow data is obtained. The data cleaning mainly comprises two types of removing default values and contradictory values, and the specific cleaning rule is as follows: removing data records with default values in four fields of inbound time, inbound site ID, outbound time and outbound site ED; and verifying the station entrance and exit time, and removing data with the station residence time being higher than 4 hours (the time can be modified by referring to the highest residence time limit of the subways in each region) or less than 2 minutes (namely the shortest one-station running time of the subways can be modified by referring to the shortest one-station running time of the subways in each region) so as to eliminate abnormal riding residence behaviors such as malicious ticket evasion, entrustment and the like. And performing data transformation on the cleaned passenger flow data of the rail transit station to obtain time series data of the rail transit station, wherein the data transformation is to identify time of different time periods and create passenger flow statistical points, and on the basis, a passenger flow time series data table of each station is generated by taking hours as a unit, and the data in the passenger flow time series data table of each station, which is the time series data of the rail transit station, is generated by taking hours as a unit.
And (4) data mining is carried out on the card swiping data of the rail transit station, namely, the structural characteristic data and the morphological characteristic data of the passenger flow of the target rail station are obtained through statistical analysis.
As shown in table 1, the structural feature data includes a mean time-average passenger flow volume of the site, an hourly passenger flow variance, a deviation degree of fluctuation of the full-day passenger flow of the site, a kurtosis degree of the full-day passenger flow of the site, and the like, wherein the time-average passenger flow volume of the site describes the total amount of the full-day passenger flow of the site; the hourly passenger flow variance describes the passenger flow equilibrium degree of each hour of the station all day; describing the symmetry of the passenger flow distribution of the whole port of the station by the station full-day passenger flow fluctuation skewness; describing the degree of steepness of the full-day passenger flow distribution of the station by the full-day passenger flow kurtosis of the station; the morphological characteristic data comprises a passenger flow peak value number, a peak-valley sudden change value and the like. The passenger flow peak number describes the number of peaks of the passenger flow of the station in the whole day, and the peak-valley sudden change value describes whether the peak-valley change of the passenger flow of the station in the whole day can be guided or not.
TABLE 1 station passenger flow characteristic index for target track station
Counting the total daily passenger flow of the target track station, calculating the passenger flow of the target track station in single working time in the whole day to obtain the station time average passenger flow of the target track station, wherein the station time average passenger flow of the target track station is expressed as
Wherein Q isGeneral assemblyThe total amount of the daily passenger flow of the target track station, and h is the daily operation duration of the target track station, and the unit is hour.
Calculating the hourly passenger flow variance of the target track station to measure the hourly passenger flow distribution condition of the target track station all day:
wherein σ2Hourly traffic variance, Q, for target track sitetIs the station traffic at the t hour.
Calculating the full-day passenger flow fluctuation skewness of the target track station to describe the symmetry of the full-day passenger flow distribution of the target track station:
wherein sigma is the standard deviation of the passenger flow of the station in hours, and S is the fluctuation skewness of the passenger flow of the target track station in the whole day.
Calculating the full-day passenger flow kurtosis of the target track station and describing the steepness degree of the value distribution of the full-day passenger flow time sequence of the target track station:
and K is the full-day passenger flow kurtosis of the target track station, and sigma is the standard deviation of the passenger flow of the station in hours.
Counting the passenger flow peak value number F and the peak-valley sudden change value B of the target track station to measure the passenger flow fluctuation form of the target track station: and drawing an hour station entrance and exit passenger flow distribution form curve of the target track station by taking time as an abscissa (unit is hour) and the hour station entrance and exit passenger flow as an ordinate, and counting the peak value and the peak-valley sudden change value of the passenger flow.
And step S280, constructing a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model by taking the site characteristic quantization index as a self-variable set and taking the site passenger flow characteristic index as a factor variable set.
In one embodiment, the step of constructing the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by taking the site characteristic quantization index as a self-variable set and taking the site passenger flow characteristic index as a factor variable set comprises the following steps:
analyzing the correlation between the dependent variable set and the independent variable set by taking the site characteristic quantization index as the independent variable set and taking the site passenger flow characteristic index as the dependent variable set to obtain a correlation analysis result; according to the correlation analysis result, a passenger flow basic prediction model, a passenger flow total amount trend change model and a passenger flow fluctuation form prediction model are constructed; and fusing the passenger flow basic prediction model and the passenger flow total amount trend change model to obtain the passenger flow total amount prediction model.
Wherein, the structural characteristic data and the morphological characteristic data are taken as a dependent variable set DV, wherein, DV1Is hjQj，DV2-7Station time average passenger flow of target track station
Variance of hourly passenger flow σ2The system comprises a station full-day passenger flow fluctuation skewness S, a station full-day passenger flow kurtosis K, a passenger flow peak number F and a peak-valley sudden change value B. The attraction range activity index P, the inter-station accessibility index, the station domain accessibility index and the bus transfer index T of the target track station are calculatedbusShared bicycle transfer index TbiMotor vehicle parking point index TpSet IV as an autovariable.
And carrying out independent variable correlation analysis according to the dependent variable set DV and the independent variable set IV, wherein the independent variable correlation analysis comprises two parts, namely: analyzing the quantitative index (independent variable IV set) of the site characteristics and the basic total amount (h) of the site passenger flowjQj) The correlation of (c); and analyzing the correlation between the station characteristic quantitative indexes (independent variable IV sets) and the station passenger flow characteristic indexes (DV) to obtain a correlation analysis result. According to the correlation analysis result, based on the regression hypothesis, a passenger flow basic prediction model, a passenger flow total amount trend change model and a passenger flow fluctuation form prediction model are constructed; fusing the passenger flow basic prediction model and the passenger flow total amount trend change model to obtain a passenger flow total amount prediction model; the input of the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model is the corresponding characteristic quantization index of each station in the independent variable set IV, and the output is the corresponding station passenger flow characteristic index in the dependent variable set.
The constructed passenger flow basic prediction model is as follows:
wherein h isjQj(0)And (3) obtaining a passenger flow basic total amount prediction result of the target track station j, wherein epsilon is a regression intercept, A is a regression coefficient, and IV is an autovariate set.
Considering the increase of the total amount of rail passenger flow caused by the increase of urban population, the average GDP growth rate of five years in the city is set as an average growth coefficient, and the trend change model of the total amount of passenger flow is as follows:
wherein the content of the first and second substances,
t represents the difference between the year of new line operation and the year of prediction of new line access passenger flow,
and predicting the result of the total passenger flow of the target track station j by considering the growth rate.
Comprehensively considering the passenger flow basic prediction model and the passenger flow total amount trend change model to obtain a final passenger flow total amount prediction model:
wherein Q isjAnd predicting the result of the total passenger flow of the target track station.
The passenger flow fluctuation form prediction model is constructed as follows:
DVj＝AqpIVp+ε
wherein, DVjIs the result of the prediction of the traffic fluctuation pattern of the target track station j, ApqIndependent variable IV with number p in linear regression equation for dependent variable with number qpCoefficient of (IV)pIs an argument with sequence number p.
And step S300, calibrating parameters of the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by adopting a regression analysis method, and determining the calibrated parameters of the prediction model.
The method comprises the steps of carrying out corresponding parameter calibration on a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model by adopting a regression analysis method, wherein an autovariate set used for training is a site characteristic quantization index of an existing track line network corresponding to historical passenger flow data. And the factor set used for training is the station passenger flow characteristics corresponding to the historical data of the existing track network. The method specifically comprises the following two parts: for independent variable set IV and dependent variable DV1To carry outAnd performing multivariate linear regression analysis, and calibrating parameters of the passenger flow total quantity prediction model. For independent variable set IV and dependent variable DV2-7Performing multiple linear regression analysis, and calibrating parameters of the passenger flow fluctuation form prediction model, wherein the parameters comprise a regression coefficient epsilon, a regression intercept A and an average growth coefficient
The method is characterized in that the method is calibrated by adopting a regression analysis method through the existing rail transit station characteristic data, rail transit station passenger flow data and urban social and economic data.
And step S320, predicting by the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model according to the parameters calibrated by the prediction models to obtain the passenger flow total amount and the passenger flow fluctuation form of the target track station.
And the prediction set independent variable is a site characteristic quantization index after the new line is put into operation. The factor set used for prediction is the total passenger flow amount and the passenger flow fluctuation form of the target track station after the new line is put into operation.
According to the data-driven rail transit new line access passenger flow prediction method, rail transit station characteristic data and rail transit station passenger flow data of a target rail transit station are obtained; carrying out quantization processing on the characteristic data of the rail transit station to obtain station characteristic quantization indexes of a target rail station; carrying out statistical analysis on the passenger flow data of the rail transit station to obtain a station passenger flow characteristic index of a target rail station; taking a site characteristic quantization index as an autovariable set, taking a site passenger flow characteristic index as a cause variable set, and constructing a passenger flow total amount prediction model and a passenger flow fluctuation form prediction model; carrying out parameter calibration on the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by adopting a regression analysis method, and determining the calibrated parameters of the prediction model; the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model are used for predicting according to the parameters calibrated by the prediction model to obtain the passenger flow total amount and the fluctuation form of the target railway station, complex data investigation is not needed, the operability is higher, the station passenger flow total amount and the fluctuation form are predicted according to the influence factors of the railway traffic passenger flow rule based on the historical data of the existing station, and the method has stronger theoretical basis and higher precision.
It should be understood that, although the steps in the flowchart of fig. 1 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a portion of the steps in fig. 1 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A data-driven rail transit new line access passenger flow prediction method is characterized by comprising the following steps:
acquiring rail transit station characteristic data and rail transit station passenger flow data of a target rail station;
quantizing the rail transit station characteristic data to obtain a station characteristic quantization index of the target rail transit station;
carrying out statistical analysis on the passenger flow data of the rail transit station to obtain a station passenger flow characteristic index of the target rail station;
taking the station characteristic quantitative index as a self-variable set, and taking the station passenger flow characteristic index as a cause variable set to construct a passenger flow total prediction model and a passenger flow fluctuation form prediction model;
carrying out parameter calibration on the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model by adopting a regression analysis method, and determining a parameter calibrated by the prediction model;
and the passenger flow total amount prediction model and the passenger flow fluctuation form prediction model predict according to the parameters calibrated by the prediction model to obtain the passenger flow total amount and the passenger flow fluctuation form of the target track station.
2. The method according to claim 1, wherein the step of performing quantization processing on the rail transit station characteristic data to obtain a station characteristic quantization index of the target rail transit station comprises:
quantifying according to the land utilization related data in the rail transit station characteristic data to obtain an attraction range activity index of the target rail transit station;
quantifying according to station continuing design related data in the rail transit station characteristic data to obtain a bus transfer index, a shared single bus transfer index and a motor vehicle parking point index of the target rail transit station;
quantifying according to station reachability relevant data in the rail transit station characteristic data to obtain an inter-station reachability index and a station domain reachability index of the target rail station;
the attraction range activity index, the bus transfer index, the shared bicycle transfer index, the motor vehicle parking point index, the inter-station reachability index and the station domain reachability index form the station characteristic quantization index.
3. The method according to claim 2, wherein the formula for quantifying according to land use related data in the rail transit station characteristic data is:
wherein, the influence range of the target track site is totally k POIs, PjIs the attraction range activity of the target track station j, NmjIs the number of m-th POIs within the influence range of the target track site j, AmjIs the rate of travel, Square, of POIs of class m within the influence of target track site jjIs the area of influence of the target track station j.
4. The method according to claim 2, wherein the step of obtaining an inter-station reachability index and a station domain reachability index by performing quantization based on the station reachability-related data in the rail transit station characteristic data includes:
acquiring a track traffic line network diagram and a street network from the station reachability related data in the track traffic station characteristic data;
converting the track traffic line network diagram into a station topological graph taking the target track station as a core by performing convex space segmentation on the track traffic line network diagram;
performing spatial morphological analysis based on the station topological graph to obtain the integration level and the selectivity of the target track station;
carrying out weighted average on the integration degree and the selectivity degree to obtain an inter-station reachability index of the target track station;
abstracting the street network into a space system formed by axes based on an axis segmentation method;
calculating a global integration average value of streets directly connected with the target track station and a local integration average value of 600-meter radius of all streets in the target track station domain based on an axis relation of a space system to obtain the global integration average value and the local integration average value;
and carrying out weighted average on the global integration average value and the local integration average value to obtain the station domain reachability index of the target track station.
5. The method of claim 1, wherein the step of performing statistical analysis on the rail transit station passenger flow data to obtain the station passenger flow characteristic index of the target rail station comprises:
carrying out data cleaning and data transformation on the rail transit station passenger flow data to obtain rail transit station time sequence data;
performing statistical analysis according to the track station time sequence data to obtain structural feature data and morphological feature data of the target track station;
and the structural characteristic data and the morphological characteristic data are used as station passenger flow characteristic indexes.
6. The method according to claim 1, wherein the step of constructing the total passenger flow prediction model and the passenger flow fluctuation form prediction model by using the site characteristic quantitative index as a self-variable set and using the site passenger flow characteristic index as a dependent variable set comprises:
analyzing the correlation between the dependent variable set and the independent variable set by taking the station characteristic quantization index as an independent variable set and taking the station passenger flow characteristic index as a dependent variable set to obtain a correlation analysis result;
according to the correlation analysis result, a passenger flow basic prediction model, a passenger flow total amount trend change model and a passenger flow fluctuation form prediction model are constructed;
and fusing the passenger flow basic prediction model and the passenger flow total amount trend change model to obtain a passenger flow total amount prediction model.
7. The method of claim 6, wherein the basic predictive model of passenger flow is:
wherein h isjQj(0)And (3) obtaining a passenger flow basic total amount prediction result of the target track station j, wherein epsilon is a regression intercept, A is a regression coefficient, and IV is an autovariate set.
8. The method of claim 7, wherein the total passenger flow trend model is:
wherein the content of the first and second substances,
t represents the difference between the year of new line operation and the year of prediction of new line access passenger flow,
and predicting the result of the total passenger flow of the target track station j by considering the growth rate.
9. The method of claim 8, wherein the total passenger flow prediction model is:
wherein Q isjAnd predicting the result of the total passenger flow of the target track station.
10. The method of claim 6, wherein the passenger flow fluctuation pattern prediction model is:
DVj＝AqpIVp+ε
wherein, DVjIs the result of the prediction of the traffic fluctuation pattern of the target track station j, ApqIndependent variable IV with number p in linear regression equation for dependent variable with number qpCoefficient of (IV)pIs an argument with sequence number p.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110312209.1A CN113077079A (en) | 2021-03-24 | 2021-03-24 | Data-driven rail transit new line access passenger flow prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110312209.1A CN113077079A (en) | 2021-03-24 | 2021-03-24 | Data-driven rail transit new line access passenger flow prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113077079A true CN113077079A (en) | 2021-07-06 |
Family
ID=76613867
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110312209.1A Pending CN113077079A (en) | 2021-03-24 | 2021-03-24 | Data-driven rail transit new line access passenger flow prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113077079A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116306049A (en) * | 2023-05-24 | 2023-06-23 | 北京城建交通设计研究院有限公司 | Rail transit connection prediction method, system, equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208034A (en) * | 2013-03-22 | 2013-07-17 | 北京交通大学 | Rail transit passenger flow distribution predicting model building method and predicting method |
| WO2016045195A1 (en) * | 2014-09-22 | 2016-03-31 | 北京交通大学 | Passenger flow estimation method for urban rail network |
-
2021
- 2021-03-24 CN CN202110312209.1A patent/CN113077079A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208034A (en) * | 2013-03-22 | 2013-07-17 | 北京交通大学 | Rail transit passenger flow distribution predicting model building method and predicting method |
| WO2016045195A1 (en) * | 2014-09-22 | 2016-03-31 | 北京交通大学 | Passenger flow estimation method for urban rail network |
Non-Patent Citations (1)
| Title |
|---|
| 巫诚诚: "基于大数据的多维度轨道站点客流直接预测模型研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, no. 4, 15 April 2018 (2018-04-15), pages 23 - 60 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116306049A (en) * | 2023-05-24 | 2023-06-23 | 北京城建交通设计研究院有限公司 | Rail transit connection prediction method, system, equipment and storage medium |
| CN116306049B (en) * | 2023-05-24 | 2023-08-15 | 北京城建交通设计研究院有限公司 | Rail transit connection prediction method, system, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | Customer preference and station network in the London bike-share system | |
| Xia et al. | Ranking station importance with human mobility patterns using subway network datasets | |
| Chen et al. | Surrogate‐based optimization of expensive‐to‐evaluate objective for optimal highway toll charges in transportation network | |
| Nocera et al. | Assessing carbon emissions from road transport through traffic flow estimators | |
| Kato et al. | Comparative analysis of transit assignment: evidence from urban railway system in the Tokyo Metropolitan Area | |
| Tiedemann et al. | Concept of a data thread based parking space occupancy prediction in a berlin pilot region | |
| CN112419711B (en) | Closed parking lot parking demand prediction method based on improved GMDH algorithm | |
| CN116681176A (en) | Traffic flow prediction method based on clustering and heterogeneous graph neural network | |
| CN115204477A (en) | Bicycle flow prediction method of context awareness graph recursive network | |
| Deng et al. | Heterogenous Trip Distance‐Based Route Choice Behavior Analysis Using Real‐World Large‐Scale Taxi Trajectory Data | |
| Mantouka et al. | Deep survival analysis of searching for on-street parking in urban areas | |
| Wang et al. | STLoyal: A spatio-temporal loyalty-based model for subway passenger flow prediction | |
| CN113821547B (en) | Rapid and efficient short-time prediction method, system and storage medium for occupancy of parking lot | |
| Zhu et al. | XGBoost‐Based Travel Time Prediction between Bus Stations and Analysis of Influencing Factors | |
| CN113077079A (en) | Data-driven rail transit new line access passenger flow prediction method | |
| Hofer et al. | Including traffic jam avoidance in an agent-based network model | |
| CN111008730B (en) | Crowd concentration prediction model construction method and device based on urban space structure | |
| Xing et al. | Recommendation of urban vehicle driving routes under traffic congestion: A traffic congestion regulation method considering road network equilibrium | |
| Li et al. | Data-driven traffic congestion patterns analysis: A case of Beijing | |
| CN115640756A (en) | Parking demand prediction model and method based on multi-source data and application thereof | |
| Poliziani et al. | Supply and demand analysis of a free floating bike sharing system | |
| CN115146840A (en) | Data-driven rail transit new line access passenger flow prediction method | |
| Xia et al. | Deeprailway: a deep learning system for forecasting railway traffic | |
| Vankdoth et al. | Deep intelligent transportation system for travel time estimation on spatio-temporal data | |
| Slik et al. | Transaction-driven mobility analysis for travel mode choices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |