A method for establishing an abstract grid road network based on a quadtree and vehicle GPS trajectories

Abstract

The application discloses a kind of based on four-tree and vehicle GPS track abstract grid road network establishment method, comprising the following steps: (1) according to GPS data logical relationship feature selection four-tree data structure;(2) using spatial connection operation will GPS data and actual road network data store in four-tree data structure;(3) using the GPS data embedding algorithm based on four-tree, according to taxi GPS data division basic grid unit;(4) according to road network topological structure, determine whether the connecting arc between each grid unit exists;(5) according to the shortest path distance between the road network nodes in each grid unit, the connecting arc weight of each grid unit is calculated by average;(6) establish abstract grid road network.The application can provide certain auxiliary reference for empty taxi path planning, reduce its empty rate.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of vehicle path planning, and particularly relates to a method for establishing an abstract grid road network based on a quadtree and a vehicle GPS trajectory. BACKGROUND

[0002] With the rapid development of cities, the travel demand of urban residents is also rapidly increasing. As an important part of the urban transportation system, taxis play an incomparable advantage in providing travel services for residents because of their unique service characteristics. At present, taxis in China are mainly divided into cruising taxis and pre-booking taxis, among which cruising taxis account for 70%, bringing great convenience to urban residents. However, there are many problems in the process of taxi service, the most important of which is the cruising problem of cruising taxis. According to the investigation of the actual situation, at present, a large part of cruising taxis is in a state of aimless cruising for a long time, and the cruising rate of taxis in the country is too high. Taxi cruising mainly brings the following effects:

[0003] 1. Restricting the long-term income of taxi drivers: Taxi drivers will not generate any income during the cruising process, and when the cruising time and distance are too long, the time cost and fuel cost of the driver will also increase, and the passenger income will be greatly reduced, thereby reducing the long-term income of the driver.

[0004] 2. Resource waste and environmental pollution: During the cruising process of taxis, no passenger travel service is provided, which not only wastes taxi resources, but also occupies more road resources. At present, many taxis in China are fuel consumption vehicles, and the cruising process will also increase unnecessary greenhouse gas emissions, causing environmental pollution.

[0005] 3. Leading to traffic congestion and even causing safety problems: In order to find passengers, taxi drivers generally go to areas with large passenger flow when driving the cruising taxi, in order to improve their passenger pickup success rate. This situation will generally cause regional traffic congestion and intensify the conflict between people and vehicles. In addition, the driver's attention is more focused on the road during the cruising process, which may cause traffic accidents due to distraction.

[0006] The grid-based abstract processing of the road network is to reshape the actual road network composed of road nodes and road sections into an abstract road network composed of several grid units through a certain processing method, and various data in the original actual road network are embedded in the abstract road network. Grid-based abstract road network processing can effectively reduce the scale of the state space and has significant advantages in studying low-dimensional data. There are some studies on abstract grid road networks in urban empty taxi path planning, but there are some defects: (1) Most current studies do not use reasonable methods from the perspective of data to evenly divide the study area when embedding the abstract road network; (2) The distribution of grid data in some abstract grid road networks is unreasonable, and the data sparsity is unreasonable, which will affect the subsequent solution speed; (3) Based on the existing road network modeling, many topological structures of the actual road network are discarded, resulting in limited information contained in the vehicle during the modeling process and not fully fitting the actual road network. Summary of the Invention

[0007] In order to solve the technical problems mentioned in the above background technology, the present invention proposes a method for establishing an abstract grid road network based on a quadtree and vehicle GPS trajectories.

[0008] In order to achieve the above technical objectives, the technical solution of the present invention is:

[0009] A method for establishing an abstract grid road network based on a quadtree and vehicle GPS trajectories includes the following steps:

[0010] S1. Selecting a quadtree data structure based on the logical relationship characteristics of GPS data;

[0011] S2. Use spatial join operation to store the vehicle GPS data and the actual road network data into the quadtree data structure of step S1;

[0012] S3. Using a quadtree-based GPS data embedding algorithm, the study area is divided into n basic grid cells B = {1, 2, ..., n} according to the distribution of vehicle GPS data points in the actual road network;

[0013] S4. Determine whether there are connecting arcs between each grid unit according to the road network topology. ii ;

[0014] S5. Calculate the average connection arc weight s of each grid cell based on the shortest path distance between the road network nodes in each grid cell. ij ;

[0015] S6. Output the final abstract grid network G = {B, L, S}, where L = {l ij |i, j∈B} represents the set of connecting arcs between each grid, S={sij | ij ∈L} represents a set of connection arc weights between grids.

[0016] Preferably, step S1 specifically comprises the following steps:

[0017] S11, the logical structure of the data obtained contains the logical relationship between the data, including belonging to a set, one-to-one, one-to-many and many-to-one, corresponding to set, linear structure, tree structure and graph structure respectively;

[0018] S12, the one-to-many logical relationship in step S11 and the driving trajectory formed between the GPS data points are taken as features, and a quadtree data structure based on point elements and line elements is selected.

[0019] Preferably, step S2 specifically comprises the following steps:

[0020] S21, selecting a research area, obtaining the shortest path between road nodes through map software, and storing the shortest path result in the form of a matrix;

[0021] S22, visualizing a plurality of continuous vehicle GPS data based on actual road conditions to form a driving trajectory, the driving trajectory existing in the form of a line element;

[0022] S23, using spatial connection operation, connecting different data tables into a new data table according to the spatial position relationship between vehicle GPS trajectories, the new data table including GPS data with geographic information and actual road network data, and storing the new data table into a quadtree data structure to retain the spatial position information of the data.

[0023] Preferably, in the step S21, it specifically comprises the following steps:

[0024] S211, according to the city development status, economic indicators and heat map as reference, selecting a city with GDP greater than one trillion, urbanization rate higher than 80%, more than 3 million migrant workers, and population density greater than 60 people / hm2 in the urban center area during morning and evening peak periods as the research area;

[0025] S212, using open map software to obtain the real road network in the research area, which contains road nodes and road segment lengths in the area, using the shortest path algorithm to solve the shortest path distance between nodes, and storing the shortest path distance in the form of a matrix.

[0026] Preferably, in the step S3, the GPS data embedding algorithm based on quadtree specifically comprises the following steps:

[0027] S311, input the research area, GPS driving track, node capacity n and pre-pruning requirement;

[0028] S312, initialize the quadtree data structure, parent node list and leaf node list;

[0029] S313, use spatial connection for each node in the parent node list, and embed the GPS driving track in the corresponding area in the node;

[0030] S314, judge each node in the parent node list, if the number of driving tracks in the node is greater than the node capacity and the minimum side length of the corresponding area is greater than the pre-pruning requirement, evenly split the node, and remove it from the parent node list, and add the four child nodes obtained by splitting to the parent node list; otherwise, add the node to the leaf node list, and remove it from the parent node list;

[0031] S315, output the finally obtained leaf node list.

[0032] Preferably, the step S3 specifically comprises the following steps:

[0033] S321, set the node capacity n, n is between 1 / 10000 and 1 / 15000 of the total GPS points;

[0034] S322, set the pre-pruning requirement C0, and set the minimum side length of the square sub-area as C0;

[0035] S323, divide the research area, each area corresponds to a node in the quadtree, when the number of GPS tracks in the area is greater than the capacity n and the minimum side length of the area is greater than C0, evenly divide the area into four square sub-areas; if the number of GPS tracks in the area is less than the capacity n or the minimum side length of the area is less than C0, stop dividing the area;

[0036] S324, continuously divide the area according to the division rule, when a sub-area does not meet the division requirement, the node corresponding to the sub-area is set as a leaf node;

[0037] S325, repeat the above steps, when all areas are divided and no new sub-nodes are generated, the final road network grid is formed, and the division is completed.

[0038] Further, the pre-pruning requirement C0 in step S322 is expressed as follows:

[0039]

[0040] In the formula, s i represents the distance between the ith pair of two consecutive GPS track points, defined as a "distance pair";

[0041] n s denotes the total number of "distance pairs" in the data set; denotes the ratio of the number of "distance pairs" with a length less than C0to the total number of "distance pairs".

[0042] Preferably, step S4 specifically comprises the following steps:

[0043] S41, preset the connection arc between the i-grid and the j-grid as (1-δ)l ij , select the road nodes covered by the i-grid and the j-grid, and determine the reachability of the nodes between the grids according to the actual road network topology;

[0044] S42, if there is reachability, then δ=0, and if there is no reachability, then δ=1;

[0045] S43, output the final connection arc l between each grid unit ij .

[0046] Preferably, the step S5 specifically comprises the following steps:

[0047] S51, use the start and end point labels to represent the shortest path distance between the actual road network nodes, and the formula is expressed as follows:

[0048] MN=d i , N∈B j

[0049] In the formula, B i denotes the set of road network nodes contained in the i-grid; B j denotes the set of road network nodes contained in the j-grid; and d denotes the shortest distance between two nodes of the actual road network.

[0050] S52, calculate the connection arc weight of each grid unit by averaging, and the formula is expressed as follows:

[0051]

[0052] In the formula, s ij denotes the connection arc weight between the i-grid and the j-grid; n denotes the actual road section number between the i-grid and the j-grid; and ∑MN denotes the sum of the shortest path distances between all nodes between the i-grid and the j-grid.

[0053] The above technical scheme has the following beneficial effects:

[0054] The application selects the quadtree data structure which is matched with the characteristics of GPS data track points and area division process when selecting data structure, sets reasonable capacity range according to GPS data track, and stipulates the minimum edge length of each area, which uniformly carries out grid division to the research area while considering data rationality, and has scientificity and rationality.

[0055] In the process of dividing the grid, the application carries out regular and uniform division according to the data amount in the grid, so that the data amount contained in each grid is as small as possible, effectively solves the problem of data sparseness, and improves the speed of subsequent problem solving.

[0056] In the process of embedding data, the application uses spatial connection operation, embeds GPS data with spatial information into the quadtree data according to actual GPS track data, and judges the existence of the connecting arc between the grids according to the topological structure of the actual road network, so that the finally formed abstract road network grid contains more actual road network topological details, and the vehicle state information contained in the modeling is more abundant. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is a method flowchart of the application;

[0058] Figure 2 is a spatial distribution diagram of the all-day GPS track data of the application;

[0059] Figure 3 is a Shenzhen road network diagram of the application;

[0060] Figure 4 is a Shenzhen road network section length distribution diagram of the application;

[0061] Figure 5 is an actual node multi-source shortest path matrix diagram of the application;

[0062] Figure 6 is an abstract road network division result diagram of the application in different time periods;

[0063] Figure 7 is a micro-grid road network diagram of the application. DETAILED DESCRIPTION

[0064] The technical solutions of the application will be described in detail below with reference to the drawings.

[0065] The application discloses an abstract grid road network establishment method based on a quadtree and vehicle GPS track, and the specific steps are as shown in Figure 1 .

[0066] According to the logical relationship characteristics of GPS data, the quadtree data structure is selected. The logical structure of data mainly includes set, linear structure, tree structure and graph structure, and the corresponding logical relationship is same set, many-to-one, one-to-many and many-to-many. The process of dividing the grid is to divide a region into multiple regions, which conforms to the logical relationship of one-to-many. This time, 797 taxi GPS data of Shenzhen City in October 2013 is selected, and the spatial distribution of the data is shown in Figure 2 . It has obvious point elements and line elements. Therefore, the quadtree data structure based on point elements and line elements is selected.

[0067] The GDP in 2022 was 32387.68 billion yuan, and it is the only city in China with a urbanization rate of 100%. The number of migrant workers in 2020 exceeded 4.9 million, and has been increasing for many years. The population density in the central area during the morning and evening peak hours is more than 60 people / hm 2 , which meets the selection conditions of the study area. According to the actual situation, the study area is determined to be 113°46′ east longitude to 114°37′ east longitude and 22°27′ north latitude to 22°52′ north latitude. The motor vehicle road network data of Shenzhen City is obtained by using the OpenStreetMap platform, and the osm2gmns plug-in is used to eliminate sub-circles with less than 5000 nodes, and the processed motor vehicle road network data is shown in Figure 3 . The length distribution of the road section is shown in Figure 4 . The shortest path algorithm is used to solve the shortest path between the nodes in the actual case and store it in matrix form, as shown in Figure 5 . Next, according to the "time stamp" and "passenger state" information of GPS data, the data is cleaned, and the processed GPS data is integrated to obtain the pick-up and drop-off points of the vehicle, as shown in Figure 6 . Then, according to the actual road network information, the GPS data and road information are stored in the quadtree data structure by using spatial connection.

[0068] The GPS data embedding algorithm based on quadtree is used to divide the study area. The node capacity n is 10, which is 1 / 13000 of the total number of trajectory points. According to the analysis of GPS trajectory points, more than 90% of the two consecutive trajectory points have a distance of less than 600m throughout the day, so C0 is set to 600m according to the formula . Combined with the spatial connection result of the empty trajectory, the GPS data embedding algorithm based on quadtree is used to divide the study area into multiple road network grids at different times, and the result is shown in Figure 7 .

[0069] Determine whether there is a connection arc between each grid unit. Because the number of divided grids is too large, it is decided to use a micro-grid road network to explain and illustrate, asFigure 7 , the circular node is the actual road network node. First, the connection arcs of the grids 1 to 13 are initialized as (1-δ)l ij , and then the connection of the actual road network node is used to determine whether there is a connection arc between two grids. If there is, such as the connection between the grid 1 and the grid 10, δ is set to 0, and the connection arc l 1，10 is output; if there is not, such as the connection between the grid 2 and the grid 7, δ is set to 1, and 0 is output.

[0070] According to the determination of the shortest path between each node, the weight of the connection arc between each grid is calculated. For example, Figure 7 , the weight s 1，6 between the grid 1 and the grid 6 is calculated. First, the distance between the grids in the actual road network is determined, then AH=9, AM=9, AI=19, BH=7, BM=7, BI=17, and the weight is calculated. Finally, the weight s ij between each grid is calculated according to the formula.

[0071] According to the divided road network grid B in front, the determined connection arc L between the grids, and the calculated connection arc weight S, the abstract grid road network G={B, L, S} is output. The empty taxi driver moves according to the moving track of the past GPS data, and moves to pick up passengers through the grid road network, which greatly reduces the empty driving rate.

[0072] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0073] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices produce a machine that implements the flow Figure 1 flow or multiple flows and / or blocksFigure 1 means for performing the function specified by the block or blocks.

[0074] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 flow or flows and / or blocks Figure 1 means for performing the function specified by the block or blocks.

[0075] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions that are executed on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 flow or flows and / or blocks Figure 1 steps for performing the function specified by the block or blocks.

[0076] Although preferred embodiments of the application have been described herein, changes and modifications can be suggested to one skilled in the art and are intended to be encompassed within the scope of the application as set forth in the appended claims. It is the intent, therefore, to be limited only as

[0077] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover the modifications and variations of this application provided they come within the scope of the appended claims and their equivalents.

Claims

1. A method for establishing an abstract grid road network based on quadtree and vehicle GPS trajectory, characterized in that: The following steps are involved: S1. Selecting a quadtree data structure based on the logical relationship characteristics of GPS data; S2. Use spatial join operation to store the vehicle GPS data and the actual road network data into the quadtree data structure of step S1; S3. Using a quadtree-based GPS data embedding algorithm, the study area is divided into n basic grid cells B = {1, 2, ..., n} according to the distribution of vehicle GPS data points in the actual road network; S4. Determine whether there are connecting arcs between grid cells based on the road network topology. ; S5. Calculate the average connection arc weight of each grid cell based on the shortest path distance between the road network nodes in each grid cell. ; S6. Output the final abstract grid road network in, represents the set of connecting arcs between each grid, Represents the set of arc weights connecting meshes; The quadtree-based GPS data embedding algorithm in step S3 specifically includes the following steps: S311, input the study area, GPS driving trajectory, node capacity n and pre-pruning requirements; S312, initialize the quadtree data structure, parent node list and leaf node list; S313: Use spatial connection for each node in the parent node list to embed the GPS driving trajectory in the corresponding area in the node; S314. Each node in the parent node list is judged. If the number of empty trajectories of the node is greater than the node capacity and the length of the shortest side of the corresponding area is greater than the pre-pruning requirement, the node is evenly split and removed from the parent node list. The four child nodes obtained by the split are added to the parent node list. Otherwise, the node is added to the leaf node list and removed from the parent node list. S315: Output the final leaf node list; In step S3, the study area is divided into n basic grid cells B={1, 2, ..., n} according to the distribution positions of the vehicle GPS data points in the actual road network, which specifically includes the following steps: S321, set the node capacity n, n value between the total GPS points ; S322. Set pre-pruning requirements , the minimum side length of the square sub-region area is ; S323, divide the study area, each area corresponds to a node in the quadtree, when the number of GPS tracks in the area is greater than the capacity n and the shortest side length of the area is greater than When , the area is evenly divided into four square sub-areas; if the number of GPS tracks in the area is less than the capacity n or the length of the shortest side of the area is less than C0, the area division is stopped; S324, continuously dividing the region according to the division rule. When a sub-region no longer meets the division requirements, the node corresponding to the sub-region is set as a leaf node; S325. Repeat the above steps until all areas are divided and no new sub-nodes are generated, forming the final road network grid and the division is completed. The pre-pruning requirement in step S322 The formula is as follows: Where, represents the distance between two consecutive GPS track points of the ith pair, defined as "distance pair"; Indicates the total number of "distance pairs" in the dataset; Indicates length is less than The ratio of the number of "distance pairs" to the total number of "distance pairs".

2. The method for establishing an abstract grid road network based on a quadtree and vehicle GPS trajectories according to claim 1, characterized in that: Step S1 specifically includes the following steps: S11. The logical structure of the acquired data includes the logical relationships between the data, including belonging to the same set, one-to-one, one-to-many, and many-to-one, corresponding to the set, linear structure, tree structure, and graph structure respectively; S12: Taking the one-to-many logical relationship in step S11 and the driving trajectory formed between the GPS data points as features, a quadtree data structure based on point elements and line elements is selected.

3. The method for establishing an abstract grid road network based on quadtree and vehicle GPS trajectory according to claim 1, characterized in that: Step S2 specifically includes the following steps: S21. Select a study area, obtain the shortest path between road nodes using map software, and store the shortest path results in the form of a matrix; S22, visually presenting a plurality of continuous vehicle GPS data based on actual road conditions to form a driving trajectory, where the driving trajectory exists in the form of a line element; S23. Using a spatial join operation, different data tables are connected into a new data table based on the spatial position relationship between the vehicle GPS trajectories. The new data table includes GPS data with geographic information and actual road network data. The new data table is stored in a quadtree data structure to retain the spatial position information of the data.

4. The method for establishing an abstract grid road network based on a quadtree and vehicle GPS trajectories according to claim 3, characterized in that: Step S21 specifically includes the following steps: S211. Based on the current status of urban development, economic indicators, and heat maps, we selected cities with a GDP greater than one trillion yuan, an urbanization rate greater than 80%, a migrant worker population of more than 3 million, and a population density greater than 60 people / hm2 in the city center during peak hours. 2 The city was used as the study area; S212. Use open map software to obtain the actual road network in the study area, which includes the road nodes and section lengths in the area, use the shortest path algorithm to solve and obtain the shortest path distance between each node, and store the shortest path distance in the form of a matrix.

5. The method for establishing an abstract grid road network based on quadtree and vehicle GPS trajectory according to claim 1, characterized in that: Step S4 specifically includes the following steps: S41, presetting the connection arc between the i grid and the j grid to be , select the road nodes covered by grid i and grid j, and determine the accessibility of the nodes between grids according to the actual road network topology; S42. If reachability exists, then , if there is no reachability, then ; S43, output the final connection arcs between the grid cells .

6. The method for establishing an abstract grid road network based on quadtree and vehicle GPS trajectory according to claim 1, characterized in that: Step S5 specifically includes the following steps: S51. Use the start and end point numbers to represent the shortest path distance between actual road network nodes. The formula is as follows: Where, Represents the set of road network nodes contained in the i-th grid; Represents the set of road network nodes contained in the j-th grid; Indicates the shortest distance between two nodes in the actual road network; S52. Calculate the average connection arc weight of each grid cell. The formula is as follows: Where, represents the connection arc weight between grid i and grid j; represents the actual number of road sections between grid i and grid j; It represents the sum of the shortest path distances between all nodes of grid i and grid j.

Images