A dynamic park-and-ride demand prediction method and system based on travel isochrone

Abstract

The present application relates to a kind of dynamic park-and-ride demand prediction method and system based on travel isochrone, its method includes: based on travel isochrone calculation is reached by the space grid of car travel and car-subway travel accessible all the way;Search shortest path;According to the shortest path, travel time is calculated and utility function is constructed in combination with travel expense;Park-and-ride advantage range is calculated according to utility function;According to the travel demand of all car individuals, the travel origin is obtained;According to the travel origin of all car individuals and park-and-ride advantage range, basic transfer demand is calculated;Basic transfer demand is corrected;Based on the temporary parking demand proportion of P+R parking lot, combined with the corrected transfer demand, the total demand of P+R parking lot is calculated.The present application realizes the accurate prediction of park-and-ride demand and the dynamic evaluation of planning scheme, scientifically guides the planning and layout of park-and-ride facilities, and provides key support for urban traffic fine management and sustainable development.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of urban traffic planning, and particularly relates to a dynamic parking and transfer demand prediction method and system based on travel isochrone. BACKGROUND

[0002] Planning and constructing P+R transfer parking lots around rail transit stations is an important strategy to help alleviate traffic congestion and guide green travel. The scale of traditional parking and transfer facilities is mainly determined according to rail passenger flow and transfer ratio. Due to the lack of spatial dimension, insufficient dynamic adaptability, and the absence of multi-factor coupling, the current parking and transfer facility planning generally has the realistic dilemma of imbalance between supply and demand and space-time mismatch, which greatly restricts the effective functioning of transfer parking lots. Although a small number of existing studies have proposed parking and transfer demand prediction methods based on individual choice behavior analysis or network equilibrium analysis, there are still certain limitations and technical bottlenecks, mainly manifested in:

[0003] (1) "Macro-micro" fragmentation: existing methods each focus on a single scale, lack a unified framework that integrates individual behavior and network effects, leading to idealized or fragmented prediction results.

[0004] (2) Insufficient dynamic adaptability: real-time traffic conditions such as road congestion, road network capacity, and rail operation are not considered in the dynamic influence on transfer demand, making it difficult to support fine-grained facility layout.

[0005] (3) Lack of multi-modal competition modeling: there is no quantitative model of cooperation-competition between car and rail travel modes, making it impossible to scientifically assess the interception effect of transfer facilities.

[0006] Parking and transfer is essentially a multi-modal traffic network that cooperates and competes with each other. With the advancement of intelligent transportation, how to establish a parking and transfer demand prediction model that takes into account individual behavior and road network dynamics has become a key issue for the scientific planning and precise allocation of parking facilities. It is a quantitative tool for serving "public transportation first" and "low-carbon travel" policy decisions. SUMMARY

[0007] The present application provides a dynamic parking and transfer demand prediction method and system based on travel isochrone to solve at least one of the above technical problems.

[0008] The technical solution of the present application to solve the above technical problems is as follows: a dynamic parking and transfer demand prediction method based on travel isochrone, comprising:

[0009] S1, dividing a traffic network of a preset area into a plurality of spatial grids, marking all spatial grid center points, rail station points and P+R parking lots, and abstracting car travel in the traffic network into two travel modes of full car travel and car-to-rail travel, and calculating spatial grids reachable by the two travel modes of full car travel and car-to-rail travel from any spatial grid center point based on an isochrone;

[0010] S2, determining any two spatial grid center points as a travel starting point and a travel ending point based on the spatial grids reachable by the two travel modes of full car travel and car-to-rail travel from any spatial grid center point, and searching for shortest paths from the travel starting point to the travel ending point by the two travel modes of full car travel and car-to-rail travel respectively by using a path search algorithm;

[0011] S3, calculating travel times according to the shortest paths from the travel starting point to the travel ending point by the two travel modes of full car travel and car-to-rail travel, and constructing a full car travel utility function and a park-and-ride travel utility function in combination with travel fees;

[0012] S4, calculating a park-and-ride advantage range according to the full car travel utility function and the park-and-ride travel utility function;

[0013] S5, establishing a traffic demand prediction model according to pre-acquired GIS spatial land use data, traffic facility data and traffic operation data, and predicting travel demand in combination with a traffic four-stage method to obtain travel starting points and travel ending points of all car individuals;

[0014] S6, calculating basic park-and-ride connection demand in the park-and-ride advantage range according to the travel starting points and the travel ending points of all car individuals and the park-and-ride advantage range;

[0015] S7, dynamically correcting the basic park-and-ride connection demand based on a rail transit reachable index, a road traffic reachable index and a road congestion index to obtain corrected park-and-ride connection demand;

[0016] S8, calculating total P+R parking lot demand based on a temporary parking demand proportion of the P+R parking lot and in combination with the corrected park-and-ride connection demand.

[0017] On the basis of the above technical solution, the application can be further improved as follows.

[0018] Further, in the S2, the path search algorithm is specifically a path search algorithm.

[0019] Further, in the S3, the utility function of the whole-car travel is represented as:

[0020]

[0021] wherein, is the utility function of the whole-car travel from the trip origin to the trip destination; is the peak congestion delay index, is the conversion coefficient of cost and time; is the cost of the whole-car travel from the trip origin to the trip destination; is the time of the whole-car travel from the trip origin to the trip destination, and , is the length of the shortest path of the whole-car travel from the trip origin to the trip destination, is the car moving speed.

[0022] Further, the shortest path of the car-to-rail travel from the trip origin to the trip destination includes the P+R parking lot accessible within the car access range around the trip origin, the rail station within the walking accessible range around the P+R parking lot, and the rail station within the walking accessible range around the trip destination.

[0023] In the S3, the utility function of the park-and-ride travel is represented as:

[0024]

[0025] wherein, is the utility function of the park-and-ride travel from the trip origin to the trip destination; is the time of the park-and-ride travel from the trip origin to the trip destination; is the conversion coefficient of cost and time; is the cost of the park-and-ride travel from the trip origin to the trip destination, and is the subway fare, is the parking fee of the P+R parking lot.

[0026] Further, the calculation formula of the time of the park-and-ride travel from the trip origin to the trip destination is:

[0027]

[0028] wherein, is the peak congestion delay index;​​​ is a time for reaching a P+R parking lot reachable within a car access range around the trip origin from the trip origin by car travel mode, and , is a car access distance from the trip origin to the P+R parking lot reachable within the car access range around the trip origin, is a car moving speed; is a time for reaching a rail station within a walking access range around the P+R parking lot from the P+R parking lot by walking mode; is a time for reaching a rail station within a walking access range around the trip destination from a rail station within the walking access range around the P+R parking lot by rail transit mode, and + / , is a rail line is a departure interval, is a distance between the rail station within the walking access range around the P+R parking lot and the rail station within the walking access range around the trip destination, is a rail average running speed; is a time for reaching the trip destination from the rail station within the walking access range around the trip destination by walking mode.

[0029] Further, the S4 is specifically:

[0030] based on a preset car access distance, the full-trip car travel utility function and the park-and-ride trip utility function are used to calculate a demarcation line at which the park-and-ride trip utility in a traffic network is equal to the full-trip car travel utility;

[0031] The demarcation line is extracted by using a convex hull algorithm, and a region enclosed by the extracted demarcation line is the park-and-ride advantage range.

[0032] Further, the S6 is specifically:

[0033] The trip origin of each car individual is marked.

[0034] The car individuals whose trip origins are located in the park-and-ride advantage range are searched and counted, and a basic park-and-ride access demand in the park-and-ride advantage range is obtained.

[0035] Further, in the S7, a formula for dynamically correcting the basic park-and-ride access demand is:

[0036] ;

[0037] wherein, a modified transfer demand, a basic transfer demand, a rail accessibility index, a road accessibility index, a traffic congestion index.

[0038] Further, in the S8, the calculation formula of the total demand of the P+R parking lot is:

[0039] ;

[0040] wherein, the total demand of the P+R parking lot, the modified transfer demand, a temporary parking demand ratio.

[0041] Based on the above-mentioned dynamic parking transfer demand prediction method based on travel isochrone, the application further provides a dynamic parking transfer demand prediction system based on travel isochrone.

[0042] A dynamic parking transfer demand prediction system based on travel isochrone, comprising:

[0043] A travel isochrone calculation module, which is used for dividing a traffic network of a preset area into a plurality of spatial grids, marking all spatial grid center points, rail stations and P+R parking lots, and abstracting car travel into two travel modes of full-car travel and car transfer rail travel in the traffic network, and calculating spatial grids reachable by the two travel modes of full-car travel and car transfer rail travel from any spatial grid center point based on travel isochrone;

[0044] A path search module, which is used for determining any two spatial grid center points as a travel starting point and a travel ending point based on spatial grids reachable by the two travel modes of full-car travel and car transfer rail travel from any spatial grid center point, and searching for shortest paths from the travel starting point to the travel ending point by the two travel modes of full-car travel and car transfer rail travel respectively by using a path search algorithm;

[0045] A utility function calculation module, which is used for calculating travel time according to the shortest paths from the travel starting point to the travel ending point by the two travel modes of full-car travel and car transfer rail travel, and constructing full-car travel utility function and parking transfer travel utility function in combination with travel cost;

[0046] A parking transfer advantage range determination module, which is used for calculating a parking transfer advantage range according to the full-car travel utility function and the parking transfer travel utility function;

[0047] a travel prediction module configured to establish a traffic demand prediction model according to pre-acquired GIS spatial land use data, traffic facility data and traffic operation data, and to predict travel demand in combination with a traffic four-stage method to obtain travel origins and travel destinations of all individual cars;

[0048] a transfer connection demand calculation module configured to calculate basic transfer connection demand within a parking and transfer advantage range according to the travel origins and the travel destinations of all individual cars and the parking and transfer advantage range;

[0049] a transfer connection demand correction module configured to dynamically correct the basic transfer connection demand based on a rail transit accessibility index, a road traffic accessibility index and a road congestion index to obtain corrected transfer connection demand;

[0050] a P+R parking lot total demand calculation module configured to calculate P+R parking lot total demand based on a temporary parking demand proportion of the P+R parking lot in combination with the corrected transfer connection demand.

[0051] The present application has the beneficial effect that the dynamic parking and transfer demand prediction method and system based on travel isochrone comprehensively considers influencing factors of parking and transfer behavior and establishes travel utility functions in multiple modes, calculates and delimits a transfer advantage space range based on a travel isochrone model, establishes a traffic demand prediction model by fusing multi-source data such as land use, traffic facilities and traffic operation, calculates car travel demand within a parking and transfer isochrone, dynamically corrects car transfer demand by comprehensively considering influencing factors such as rail accessibility, road accessibility and traffic congestion state to obtain corrected parking and transfer demand, and superimposes temporary parking demand proportions of P+R parking lots in different traffic locations to obtain P+R parking lot total demand; the present application fuses classical isochrone theory in the planning field and the traffic planning four-stage method, quantitatively models competitive-complementary relationships of multiple travel modes, establishes a spatiotemporal coupling isochrone dynamic boundary algorithm, realizes accurate prediction of parking and transfer demand and dynamic evaluation of planning schemes through the innovative framework of "travel isochrone+multi-scale fusion+dynamic correction", and scientifically guides parking and transfer facility planning and layout to provide key support for urban traffic fine management and sustainable development. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 a flowchart of the dynamic parking and transfer demand prediction method based on travel isochrone of the present application;

[0053] Figure 2 a schematic diagram of car travel and car transfer rail travel modes;

[0054] Figure 3This is a structural block diagram of a dynamic park-and-ride demand prediction system based on travel isochronous circles according to the present invention. Detailed Implementation

[0055] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0056] Example 1:

[0057] like Figure 1 As shown, a dynamic park-and-ride demand prediction method based on travel isochronous circumferences includes:

[0058] S1 divides the traffic network of the preset area into multiple spatial grids, marks all spatial grid center points, rail stations and P+R parking lots, and abstracts car travel in the traffic network into two modes of travel: all-car travel and car-to-rail transit travel. Based on the travel isochronous circle, it calculates the spatial grids that any spatial grid center point can reach through the two modes of travel: all-car travel and car-to-rail transit travel.

[0059] S2, based on a spatial grid that can be reached from any spatial grid center point by two modes of travel: full-journey car travel and car-to-rail transit travel, any two spatial grid center points are determined as the travel start point and travel end point, and the path search algorithm is used to search for the shortest path from the travel start point to the travel end point by the full-journey car travel and car-to-rail transit travel respectively.

[0060] S3. Calculate the travel time from the starting point to the destination via the shortest path of car travel and car-rail transit transfer, and combine the travel cost to construct the car travel utility function and the park-and-ride travel utility function.

[0061] S4. Calculate the parking and ride-hailing advantage range based on the total car travel utility function and the parking and ride-hailing travel utility function.

[0062] S5. Based on the pre-acquired GIS spatial land use data, traffic facility data and traffic operation data, a traffic demand prediction model is established, and the four-stage traffic method is combined to predict travel demand, so as to obtain the travel origin and travel destination of all individual cars.

[0063] S6. Based on the starting and ending points of all individual cars and the aforementioned parking and transfer advantage range, calculate the basic transfer and connection demand within the parking and transfer advantage range.

[0064] S7, dynamically revising the basic transfer demand based on the rail transit accessibility index, the road traffic accessibility index, and the road congestion index to obtain a revised transfer demand;

[0065] S8, calculating a total demand of the P+R parking lot based on a temporary parking demand ratio of the P+R parking lot and the revised transfer demand.

[0066] The following specifically explains each step:

[0067] Preferably, in the S1, the car travel includes two travel modes of the whole-process car travel and the car transfer rail transit travel, Figure 2 The whole-process car travel and the car transfer rail transit travel mode are shown in the schematic diagram; wherein, O point is a travel starting point, D point is a travel ending point, P point is a P+R parking lot accessible within a car transfer range around the O point, S point is a rail station within a walking accessible range around the P point, S point is a rail station within a walking accessible range around the D point; the trajectory OD represents the whole-process car travel mode; the trajectory OPSS D represents the car transfer rail transit travel mode; wherein, OP is a car driving trajectory, PS is a walking trajectory, SS is a rail transit trajectory, and S D is a walking trajectory.

[0068] The travel isochrone analysis is to take a specific point as the center, and to find the geographical space range accessible within a fixed time through different traffic modes.

[0069] Preferably, in the S2, according to the determined travel starting point and travel ending point, the shortest path between the travel starting point O and the travel ending point D in the whole-process car travel and the shortest path in the car transfer rail transit travel are searched respectively based on the walking road network and the car road network, and according to the walking moving speed and the car moving speed , by using a path search algorithm . The path search algorithm is a heuristic algorithm widely used in graph search and path planning, which is used to find the optimal path from the starting point to the target node.

[0070] Preferably, in the S3, the travel utility function considers the whole-process travel time and travel cost; since the individual may have one or more transfer behaviors in the process of displacement by rail transit, the present technical solution does not limit the number of rail transit transfers of the passenger, but only calculates the rail transit moving time length between any two stations.

[0071] The construction of the whole car travel utility function: according to the car moving speed and the shortest path from the travel starting point to the travel ending point by the whole car travel mode, the time and cost of reaching the travel ending point from the travel starting point by the car mode are calculated, and then the whole car travel utility function is constructed. The whole car travel utility function is expressed as:

[0072]

[0073] is the utility function of reaching the travel ending point from the travel starting point by the whole car travel mode; is the peak congestion delay index, is the conversion coefficient of cost and time; is the cost of reaching the travel ending point from the travel starting point by the whole car travel mode; is the time of reaching the travel ending point from the travel starting point by the whole car travel mode, and is the length of the shortest path of reaching the travel ending point from the travel starting point by the whole car travel mode, is the car moving speed.

[0074] The construction of the park-and-ride travel utility function includes:

[0075] (1) Trajectory OP travel time calculation: the car travel path search algorithm is used to search the P+R parking lot P that can be reached within the car connection range around the travel starting point O, and calculate the car travel time from the travel starting point O to the P+R parking lot P that can be reached around the travel starting point; the time of reaching the P+R parking lot from the travel starting point by the car travel mode is expressed as:

[0076]

[0077] is the time of reaching the P+R parking lot from the travel starting point by the car travel mode, is the car connection distance from the travel starting point to the P+R parking lot that can be reached around the travel starting point, according to the comparison of the investigation data of each city, take ≤12km, is the car moving speed.

[0078] ​​​​​(2) Calculation of travel time for trajectory PS: Starting from the P+R parking lot P, the walking path search algorithm is used to search for the rail station S within the walking range of the P+R parking lot P, and the travel time from the P+R parking lot to the rail station within the walking range of the P+R parking lot is calculated. .

[0079] (3) Trajectory SS Travel time calculation: Based on the arrival frequency information of the rail lines, calculate the travel time from rail station S within walking distance of the P+R parking lot to rail station within walking distance of the destination via rail line (line = 1,2,…,r). time This time includes waiting time and rail travel time;

[0080] + / ;

[0081] in, The time taken to travel from a rail station within walking distance of the P+R parking lot to a rail station within walking distance of the destination via rail transit. For the track line Departure interval The distance between a rail station within walking distance of the P+R parking lot and a rail station within walking distance of the destination. This represents the average running speed of the track.

[0082] (4) Trajectory S D. Travel time calculation: Based on rail stations within walking distance of the travel destination. Starting from point D, calculate the time to travel to the destination D. .

[0083] (5) Calculate the trajectory using OPSS D's travel time This refers to the time taken to travel from the starting point O to the destination D via P+R (Park and Ride) transfer to rail transit; and the time taken to travel from the starting point to the destination via car transfer to rail transit. Represented as:

[0084] ;

[0085] in, The time taken to travel from the starting point of the journey to the destination via a car and then rail transit. This represents the peak congestion delay index.

[0086] (6) Park-and-ride travel utility function construction: based on the whole process travel time and travel cost, the park-and-ride travel utility function is calculated; the park-and-ride travel utility function is expressed as:

[0087] ;

[0088] Wherein, is the utility function from the travel starting point to the travel ending point by car transfer rail transit travel mode; is the conversion coefficient of cost and time; is the cost from the travel starting point to the travel ending point by car transfer rail transit travel mode, and is the subway fare, is the P+R parking lot parking fee.

[0089] Preferably, the S4 is specifically:

[0090] Based on the preset car connection distance, the whole car travel utility function and the park-and-ride travel utility function are used to calculate the demarcation line when the park-and-ride travel utility in the traffic network is equal to the whole car travel utility.

[0091] The convex hull algorithm is used to extract the demarcation line, and the area enclosed by the extracted demarcation line is the park-and-ride advantage range.

[0092] Specifically, the park-and-ride advantage range, that is, the area on the side where the park-and-ride travel utility is lower than the whole car travel utility, and at the same time, a certain car connection range is met. First, when , and the independent variable , the demarcation line of park-and-ride and whole car travel is obtained; then the convex hull algorithm is used to extract the demarcation line, and according to the Graham scanning method, a point on the convex hull is found, and the points on the convex hull are found in the counterclockwise direction from the point, and the polar angle sorting is completed. The extraction of the convex hull is completed, and the enclosed area of the extracted demarcation line is the park-and-ride advantage range.

[0093] Preferably, in the S5, the GIS space land data includes land properties, land areas, volume rates, etc. of each land block, the traffic facility data includes rail station facilities, P+R parking facilities, road network facilities, etc., and the traffic operation data includes rail departure frequency, road operation speed, road congestion index, etc.

[0094] The traffic four-stage method is traffic generation, traffic distribution, traffic mode division, and traffic distribution; in the traffic distribution in the embodiment, the user equilibrium model is adopted. Preferably, the S6 is specifically:

[0095] Marking the trip origin of each car individual;

[0096] Searching and counting the car individuals whose trip origins are located in the park-and-ride advantage range, to obtain the basic park-and-ride connection demand in the park-and-ride advantage range.

[0097] Preferably, in the S7, the formula for dynamically correcting the basic park-and-ride connection demand is:

[0098] ;

[0099] Wherein, is the corrected park-and-ride connection demand, is the basic park-and-ride connection demand, is the rail accessibility index, is the road accessibility index, is the traffic congestion index.

[0100] Specifically, the present application considers the influence of rail transit accessibility, road traffic accessibility and road congestion state in different areas on park-and-ride travel, so as to dynamically correct the basic park-and-ride connection demand according to the rail transit accessibility index, the road accessibility index and the traffic congestion index.

[0101] Preferably, in the S8, the calculation formula of the total demand of the P+R parking lot is:

[0102] ;

[0103] Wherein, is the total demand of the P+R parking lot, is the corrected park-and-ride connection demand, is the temporary parking demand proportion.

[0104] Specifically, the present application considers the temporary parking demand proportion of the P+R parking lot in different traffic locations, and superimposes the park-and-ride connection demand and the temporary parking demand to obtain the total demand of the P+R parking lot.

[0105] Embodiment two:

[0106] Based on the above-mentioned dynamic park-and-ride demand prediction method based on travel isochrone, the present application further provides a dynamic park-and-ride demand prediction system based on travel isochrone.

[0107] As Figure 3 shown, a dynamic park-and-ride demand prediction system based on travel isochrone comprises:

[0108] The trip isohypse calculation module is configured to divide a traffic network of a preset area into a plurality of spatial grids, mark all spatial grid center points, rail station points and P+R parking lots, and abstract car trips in the traffic network into two trip modes of full car trips and car-to-rail trips, and calculate spatial grids reachable by the two trip modes of full car trips and car-to-rail trips from any spatial grid center point.

[0109] The path search module is configured to determine any two spatial grid center points as a trip starting point and a trip ending point based on the spatial grids reachable by the two trip modes of full car trips and car-to-rail trips from any spatial grid center point, and search for shortest paths from the trip starting point to the trip ending point by the two trip modes of full car trips and car-to-rail trips respectively by using a path search algorithm.

[0110] The utility function calculation module is configured to calculate trip times according to the shortest paths from the trip starting point to the trip ending point by the two trip modes of full car trips and car-to-rail trips, and construct full car trip utility functions and car-to-rail trip utility functions in combination with trip fees.

[0111] The P+R parking lot total demand calculation module is configured to calculate P+R parking lot total demand based on a temporary parking demand proportion of the P+R parking lot and in combination with the modified transfer and connection demand.

[0112] The trip prediction module is configured to establish a traffic demand prediction model according to previously acquired GIS spatial land use data, traffic facility data and traffic operation data, and predict trip demand in combination with a traffic four-stage method to obtain trip starting points and trip ending points of all car individuals.

[0113] The transfer and connection demand calculation module is configured to calculate basic transfer and connection demand in the P+R parking lot advantage range according to the trip starting points and trip ending points of all car individuals and the P+R parking lot advantage range.

[0114] The transfer and connection demand correction module is configured to dynamically correct the basic transfer and connection demand based on a rail transit reachable index, a road traffic reachable index and a road congestion index to obtain modified transfer and connection demand.

[0115] The P+R parking lot total demand calculation module is configured to calculate P+R parking lot total demand based on a temporary parking demand proportion of the P+R parking lot and in combination with the modified transfer and connection demand.

[0116] The specific functions of the modules in the dynamic park-and-ride demand prediction system based on the travel isochrone are described in the steps of the dynamic park-and-ride demand prediction method based on the travel isochrone, which will not be repeated here.

[0117] The dynamic park-and-ride demand prediction method and system based on the travel isochrone comprehensively consider the influencing factors of park-and-ride behaviors and establish travel utility functions in multiple modes, calculate and demarcate the transfer advantage space range based on the travel isochrone model, establish a traffic demand prediction model by fusing multiple source data such as land use, traffic facilities and traffic operation, calculate the car travel demand within the park-and-ride isochrone, dynamically correct the car transfer demand by comprehensively considering influencing factors such as rail accessibility, road accessibility and traffic congestion state, obtain the corrected park-and-ride demand, and superimpose the temporary parking demand proportion of the P+R parking lot in different traffic locations to obtain the total demand of the P+R parking lot; the present application quantitatively models the competitive-complementary relationship of multi-mode travel by fusing the classic isochrone theory in the planning field and the four-stage method of traffic planning, establishes a space-time coupled isochrone dynamic boundary algorithm, realizes the accurate prediction of park-and-ride demand and the dynamic evaluation of planning schemes through the innovative framework of "travel isochrone + multi-scale fusion + dynamic correction", scientifically guides the planning and layout of park-and-ride facilities, and provides key support for fine management and sustainable development of urban traffic.

[0118] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for predicting dynamic park-and-ride demand based on travel isochronous circles, characterized in that, include: S1 divides the traffic network of the preset area into multiple spatial grids, marks all spatial grid center points, rail stations and P+R parking lots, and abstracts car travel in the traffic network into two modes of travel: all-car travel and car-to-rail transit travel. Based on the travel isochronous circle, it calculates the spatial grids that any spatial grid center point can reach through the two modes of travel: all-car travel and car-to-rail transit travel. S2, based on a spatial grid that can be reached from any spatial grid center point by two modes of travel: full-journey car travel and car-to-rail transit travel, any two spatial grid center points are determined as the travel start point and travel end point, and the path search algorithm is used to search for the shortest path from the travel start point to the travel end point by the full-journey car travel and car-to-rail transit travel respectively. S3. Calculate the travel time from the starting point to the destination via the shortest path of car travel and car-rail transit transfer, and combine the travel cost to construct the car travel utility function and the park-and-ride travel utility function. S4. Calculate the parking and ride-hailing advantage range based on the total car travel utility function and the parking and ride-hailing travel utility function. S5. Based on the pre-acquired GIS spatial land use data, traffic facility data and traffic operation data, a traffic demand prediction model is established, and the four-stage traffic method is combined to predict travel demand, so as to obtain the travel origin and travel destination of all individual cars. S6. Based on the starting and ending points of all individual cars and the aforementioned parking and transfer advantage range, calculate the basic transfer and connection demand within the parking and transfer advantage range. S7. Based on the rail transit accessibility index, road traffic accessibility index and road congestion index, the basic transfer and connection demand is dynamically adjusted to obtain the adjusted transfer and connection demand. S8. Based on the proportion of temporary parking demand in P+R parking lots and combined with the revised transfer connection demand, calculate the total demand for P+R parking lots.

2. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, In S2, the path search algorithm is specifically as follows: Path search algorithm.

3. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, In S3, the utility function for the entire car trip is expressed as: ； in, Let be the utility function for traveling from the starting point to the destination by car throughout the journey; This refers to the peak congestion delay index. This is a conversion factor for costs and time; The cost of traveling from the starting point to the destination by car throughout the entire journey; The time taken to travel from the starting point to the destination by car throughout the entire journey, and , This refers to the length of the shortest path from the starting point to the destination using only private cars. This represents the speed at which the car is moving.

4. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, The shortest path from the starting point of the trip to the destination via car transfer to rail transit includes park and ride-on parking lots within the car transfer area around the starting point of the trip, rail stations within walking distance around the park and rail stations within walking distance around the destination. In S3, the park-and-ride travel utility function is expressed as: ； in, Let $\frac{ ... The time taken to travel from the starting point of the journey to the destination via a car and then rail transit. This is a conversion factor for costs and time; The cost of traveling from the starting point of the trip to the destination by transferring from a car to rail transit, and For the cost of taking the subway, Parking fees for P+R transfer parking lots.

5. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 4, characterized in that, The formula for calculating the time required to travel from the starting point to the destination via car and then rail transit is as follows: ； in, Peak congestion delay index; The time taken to travel from the starting point to a nearby park and ride-on parking lot accessible by car within the car pick-up area of ​​the starting point is denoted as _____. , The distance from the departure point to accessible P+R parking lots within the vicinity of the departure point is the car transfer distance. The speed of the car; The time taken to reach a rail station within walking distance of the P+R parking lot from the P+R parking lot; The time taken to travel from a rail station within walking distance of the P+R parking lot to a rail station within walking distance of the destination via rail transit. + / , For the track line Departure interval The distance between a rail station within walking distance of the P+R parking lot and a rail station within walking distance of the destination. The average running speed of the track; The time taken to reach the destination by walking from any rail station within walking distance of the destination.

6. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, Specifically, S4 is: Based on the preset car connection distance, the boundary line when the utility of parking and transfer in the transportation network is equal to the utility of the total car trip is calculated using the total car trip utility function and the parking and transfer trip utility function. The boundary line is extracted using the convex hull algorithm, and the area enclosed by the extracted boundary line is the parking and transfer advantage range.

7. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, Specifically, S6 is: Mark the origin and departure points of all individual cars; Search and count individual cars whose travel origins are located within the parking and transfer advantage range to obtain the basic transfer and connection needs within the parking and transfer advantage range.

8. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, In step S7, the formula for dynamically adjusting the basic transfer connection demand is as follows: ； in, To address the aforementioned need for improved transfer and connection services. This addresses the basic transfer and connection needs. The orbital reachability index, For road accessibility index, This is the traffic congestion index.

9. The dynamic park-and-ride demand prediction method based on travel isochronous circles according to claim 1, characterized in that, In step S8, the formula for calculating the total demand for P+R parking lots is: ； in, This represents the total demand for the P+R parking lot. To address the aforementioned need for improved transfer and connection services. This represents the proportion of temporary parking demand.

10. A dynamic park-and-ride demand prediction system based on travel isochronous circles, characterized in that, include: The travel isochronous circle calculation module is used to divide the traffic network of a preset area into multiple spatial grids, mark all spatial grid center points, rail stations and P+R parking lots, and abstract car travel in the traffic network into two travel modes: all-car travel and car-to-rail transit travel. Based on the travel isochronous circle, it calculates the spatial grids that any spatial grid center point can reach through the two travel modes: all-car travel and car-to-rail transit travel. The path search module is used to determine any two spatial grid center points as the starting point and the destination, and to use the path search algorithm to search for the shortest path from the starting point to the destination via the two modes of travel: all-car travel and car-to-rail transit travel. The utility function calculation module is used to calculate the travel time from the starting point to the destination via the shortest path of car travel and car-rail transit transfer, and to construct the utility function of car travel and parking-rail transit travel by combining the travel cost. The park-and-ride advantage range determination module is used to calculate the park-and-ride advantage range based on the total car travel utility function and the park-and-ride travel utility function. The travel forecasting module is used to build a traffic demand forecasting model based on pre-acquired GIS spatial land use data, traffic facility data, and traffic operation data, and combine it with the four-stage traffic method to predict travel demand, thereby obtaining the travel origin and destination of all individual cars. The transfer and connection demand calculation module is used to calculate the basic transfer and connection demand within the parking and transfer advantage range based on the travel origin and destination of all individual cars and the parking and transfer advantage range. The transfer connection demand correction module is used to dynamically correct the basic transfer connection demand based on the rail transit accessibility index, the road traffic accessibility index and the road congestion index, so as to obtain the corrected transfer connection demand. The P+R parking lot total demand calculation module is used to calculate the total demand for P+R parking lots based on the proportion of temporary parking demand in P+R parking lots and the adjusted transfer connection demand.

Images