A method and system for intelligently recommending routes for urban marathon tracks

Abstract

The present invention relates to a method and system for intelligently recommending routes for a city marathon track, and the method comprises: S1: establishing a city road information database and a city POI database; S2: determining the starting point and the end point of the marathon track according to the city POI database, and judging the feasibility of route planning; S3: obtaining POI data of the starting point and its surroundings, and POI data of the end point and its surroundings, and screening POI data according to the city road information database, obtaining a list of POI node network topological relationships, and constructing a POI node network; S4: starting from the starting point, using a greedy algorithm and a backtracking algorithm in the POI node network to generate multiple sub-routes to include as many POI nodes as possible to form a track route list; S5: traversing the track route list, evaluating each route, and obtaining a final track recommended route list according to the evaluation score. The method provided by the present invention solves the problems of low planning efficiency and many safety hazards caused by the current track planning mainly based on manual work.

Description

Technical Field

[0001] The present invention relates to the field of route planning, and in particular to a method and system for intelligently recommending routes for a city marathon track. Background Art

[0002] The route planning of urban marathon tracks needs to follow three planning principles: considering the difficulty of managing the planned route, considering the overall humanities and historical characteristics of the route, and considering whether the track meets the requirements of the official organization. The current method used for urban marathon track route planning is to first understand the city’s landmarks, scenic spots, historical culture, etc. through information query or the organizer, pre-plan the main point information such as the starting point, passing point, and end point of the event, and then connect the pre-planned points in series through software to form different event routes. The disadvantages are low route planning efficiency; the planned route has certain limitations, and it is easy to ignore road traffic conditions, which will bring higher urban traffic congestion problems; the planned roads may have narrow roads, which will bring safety hazards to the event. Summary of the invention

[0003] In order to solve the above technical problems, the present invention provides a method and system for intelligently recommending urban marathon track routes.

[0004] The technical solution of the present invention is: a method for intelligently recommending routes for urban marathon tracks, comprising:

[0005] Step S1: Establishing a city road information database and a city POI database;

[0006] Step S2: determining the starting point and the end point of the marathon track according to the city POI database, and judging the feasibility of route planning;

[0007] Step S3: obtaining the POI data of the starting point and its surroundings, and the POI data of the end point and its surroundings, and filtering the POI data according to the urban road information database, obtaining a POI node network topology relationship list, and constructing a POI node network;

[0008] Step S4: Starting from the starting point, a greedy algorithm is used to select the adjacent node closest to the current node as the next POI node in the POI node network until the end point is reached; a backtracking algorithm is used to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible to form a track route list.

[0009] Step S5: traverse the track route list, calculate the evaluation value of each track route from four dimensions: POI popularity value, width suitability, route accessibility index and POI density, and obtain the final track recommended route list after sorting according to the evaluation value.

[0010] Compared with the prior art, the present invention has the following advantages:

[0011] The present invention discloses an intelligent recommendation method for a city marathon track route, which solves the problems of low planning efficiency and many safety hazards caused by the current manual track planning. The present invention can ensure that the planned track passes through many scenic spots, city landmarks, colleges and universities and other areas and buildings with city characteristics, enrich the cultural connotation of the track, promote the city's image, save track planning time, reduce the safety hazards caused by track planning, and at the same time, the planned track route avoids traffic congestion sections, reducing the impact on urban traffic and citizens' travel. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 This is a flow chart of a method for intelligently recommending routes for a city marathon track in an embodiment of the present invention;

[0013] Figure 2 Schematic diagram of the relationship between POI nodes and road positions in an embodiment of the present invention;

[0014] Figure 3 A schematic diagram of a network composed of POI nodes in an embodiment of the present invention;

[0015] Figure 4 It is a structural block diagram of an intelligent recommendation system for urban marathon track routes in an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The present invention provides an intelligent recommendation method for a city marathon track route, which solves the problems of low planning efficiency and many safety hazards caused by the current track planning mainly based on manual work.

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below through specific implementations and in conjunction with the accompanying drawings.

[0018] First, in order to better understand the present invention, the terms used therein are explained:

[0019] POI: Point of interest. Each POI contains at least four basic information: name, address, category, and longitude and latitude coordinates. A POI can be a house, a shop, a community entrance, a park, or a bus stop, etc.

[0020] Embodiment 1

[0021] like Figure 1 As shown, an embodiment of the present invention provides a method for intelligently recommending routes for a city marathon track, comprising the following steps:

[0022] Step S1: Establishing a city road information database and a city POI database;

[0023] Step S2: Determine the starting point and end point of the marathon track according to the city POI database and judge the feasibility of route planning;

[0024] Step S3: obtaining POI data of the starting point and its surroundings, and POI data of the end point and its surroundings, and filtering the POI data according to the urban road information database, obtaining a POI node network topology relationship list, and constructing a POI node network;

[0025] Step S4: Starting from the starting point, a greedy algorithm is used to select the adjacent node closest to the current node as the next POI node in the POI node network until the end point is reached; a backtracking algorithm is used to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible to form a track route list.

[0026] Step S5: traverse the track route list, calculate the evaluation value of each track route from four dimensions: POI heat value, width suitability, route accessibility index, and POI density, and obtain the final track recommended route list after sorting according to the evaluation value.

[0027] In one embodiment, the above step S1: establishing a city road information database and a city POI database specifically includes:

[0028] Step S11: collecting urban road information data, and collecting urban road traffic congestion data at different frequencies according to different time periods; building an urban road information database;

[0029] Access the service provider's API interface, such as using the Amap API interface to collect data on urban road information; extract road information through matching algorithms; calculate the minimum road width information based on the number of lanes and road levels in the road information; establish an urban road information database based on the road information and the minimum road width information; and supplement the urban road information database by manually inputting road information;

[0030] Collect urban road traffic congestion data from 7:30 am to 12:00 pm from Monday to Sunday, and collect data every 10 minutes from 7:30 am to 9:30 am, and every 30 minutes from 9:30 am to 12:00 pm;

[0031] Based on the above data, an urban road information database is constructed, which includes road ID, road name, name of the city where the road is located, number of road lanes, road level, minimum road width, route direction, lane congestion level from 8 am to 12 pm from Monday to Sunday, and congested road section number; lane congestion levels are divided into unimpeded, slow moving, congested and severely congested.

[0032] Step S12: Obtain POI data of POI type "square" as the starting point and end point data sets, obtain POI data within a straight-line distance of 42.195 kilometers as the surrounding POI data set based on the starting point data set and the end point data set, and construct a city POI database.

[0033] Access the service provider's API interface, for example, use the Amap API interface to collect POI data; obtain POI data with POI type "square" as the starting data set and the end data set; and obtain POI data within a straight-line distance of 42.195 kilometers as the surrounding POI data set based on the starting data set and the end data set. Build a city POI database based on the above data sets, and you can also supplement the city POI database by manually entering POI data.

[0034] The city POI database contains POI number, POI name, POI type, POI center longitude and latitude address, POI entrance longitude and latitude information, POI address, start and end POI numbers, and distances from the start and end POIs.

[0035] In one embodiment, the above step S2: determining the starting point and end point of the marathon track according to the city POI database and judging the feasibility of route planning specifically includes:

[0036] Select the starting point POI data and the end point POI data from the city POI database. If the straight-line distance between the two is less than 21 kilometers, go to step S3, otherwise reselect.

[0037] According to the "China Marathon Management Documents", the straight-line distance between the starting and ending points of Class A1 certified events cannot exceed 50% of the total distance. That is, for a full marathon, the straight-line distance between the starting and ending points needs to be less than 21 kilometers. Therefore, it is necessary to select the starting point POI data and the ending point POI data that meet this requirement from the city POI database to determine the starting point and the end point.

[0038] In one embodiment, the above step S3: obtaining POI data of the starting point and its surroundings, and POI data of the end point and its surroundings, and filtering the POI data according to the urban road information database, obtaining a POI node network topology relationship list, and constructing a POI node network, specifically includes:

[0039] Step S31: According to the selected starting point POI data and end point POI data, obtain the surrounding POI data within 42.195 kilometers corresponding to the two from the city POI database;

[0040] Step S32: According to the requirements of the marathon event in the "China Marathon Management Document", the width of the event trail should be no less than 8 meters. Therefore, according to the urban road information database, POI data near roads with a total width of one-way lanes less than 8 meters are eliminated;

[0041] Step S33: Divide the area around each POI node into four areas according to the starting point POI data, the ending point POI data and the eliminated surrounding POI data, find other POI nodes closest to each area of ​​the POI node, establish connections with them, form a topological relationship, and obtain a POI node network topological relationship list;

[0042] Step S34: traverse the POI node network topology relationship list in sequence to obtain route information to adjacent nodes; and filter the POI node network topology relationship list according to the urban road information database to obtain a node path information list and construct a POI node network.

[0043] like Figure 2 As shown in the figure, after selecting the starting point S and the end point E of the square as the POI type, the surrounding buildings A, landmarks B, scenic spots C, and buildings D are obtained as original nodes, and the roads named R1, R2, R3, R4, R5, and R6 are obtained, as well as the intersections formed by the roads connected and named 26, 14, 16, 54, and 56. Figure 3 As shown in the figure, since building A is close to roads R2 and R3, A is converted into two POI nodes A1 and A2. Scenic spot C is only close to road R6. Figure 3 In the example, landmark B and building D are only adjacent to one road, so in Figure 3 The transformations are POI nodes B1 and D1 respectively. Thus, we can get the POI node list: S, E, A1, A2, B1, C1, D1.

[0044] Divide the area around each POI node in the POI node list into four areas, namely, upper left, upper right, lower left, and lower right; traverse the POI nodes in the POI node list, find the nearest POI nodes in the four areas of the node to form a connection relationship, and form a POI node network topology relationship list. The topology relationship constrains that each POI node has at most 6 connection relationships. For example, Figure 3As shown, POI node C1 is at the lower right of POI node A1 and is the closest to it, so POI node C1 is connected to POI node A1; because POI node C1 is at the lower right of POI node A2 and is the closest to it, POI node C1 is connected to POI node A1; because POI node A1 and POI node A2 belong to the same original node A, they do not form a connection relationship; and so on, a list of POI node network topology relationships is obtained.

[0045] Finally, the POI node network topology relationship list is screened according to the urban road information database, and the road sections with traffic congestion are deleted to obtain the node path information list and construct the final POI node network, such as Figure 3 shown.

[0046] In one embodiment, the above step S4: starting from the starting point, using a greedy algorithm to select the nearest neighboring node to the current node as the next POI node in the POI node network until the end point is reached; using a backtracking algorithm to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible, and finally forming a track route list, specifically including:

[0047] Use the greedy algorithm to start from the starting point, select the POI node that has not been traversed by the current route and is closest to the current node as the next POI node, and record the distance between the POI nodes that have been traversed until the last POI node is the end point to form a track route record; when the distance of the track route is greater than 42.695 kilometers or less than 41.695 kilometers, use the backtracking algorithm to backtrack for each track route record, and use the greedy algorithm to select the route from the end point to form a sub-route and save the fallback status; fall back in sequence according to the fallback status until it falls back to the starting point; select track routes with a distance of 41.695 to 42.695 kilometers to form a list of track routes.

[0048] The principle of marathon track route planning is to select a route with a large number of POI nodes from a given starting POI node to a given ending POI node and satisfying the route length of about 49.125 kilometers. Figure 3When planning the track route for the POI node network shown in the figure, starting from the starting point S, according to the greedy algorithm, the nearest POI node A2 is found as the next node. From A2, the end point E can be reached directly, forming the SA2E route. If the length of the SA2E route is less than 41.695 kilometers, this route will not be retained. Therefore, the next possible route is found from A2, which can go from A2 to C1, then from C1 to D1, then from D1 to B1, and finally from B1 to the end point E, forming the SA2C1D1B1E route. If the route length is greater than 42.695 kilometers, this route cannot be retained; however, the backtracking algorithm can be used to find a route closer to 42.195 kilometers. In the SA2C1D1B1E route, you can trace back from the end point E to B1. Since B1 has no other adjacent nodes that have not been visited, you can trace back from B1 to D1. Since D1 has no other adjacent nodes that have not been visited, you can trace back from D1 to C1. At this time, C1 can go to B1 that has not been visited, and finally go from B1 to the end point E, forming the SA2C1B1E route. If the total distance of this route is about 42.195 kilometers, then the route is retained; then, since C1 has no adjacent nodes that have not been visited, it traces back to A2. A2 also has no adjacent nodes that have not been visited, so it traces back to S. At this time, S can also start from A1 and follow the same steps to the end point E, forming the SA1C1D1B1E, SA1C1B1E, and SA1C1A2E routes. If the lengths of these three routes are between 41.695 kilometers and 42.695 kilometers, they are all retained. Thus, a track route list L[SA2C1B1E, SA1C1D1B1E, SA1C1B1E, SA1C1A2E] is formed.

[0049] In one embodiment, the above step S5: traverses the track route list, calculates the evaluation value of each track route from four dimensions: POI popularity value, width suitability, route accessibility index, and POI density, and obtains the final track recommended route list after sorting according to the evaluation value, which specifically includes:

[0050] The evaluation value calculation formula (1) is as follows:

[0051]

[0052] Among them, Score is the final evaluation score of the track route; S poi is the score of the track route POI heat value dimension, with a full score of 100, as shown in formula (2):

[0053]

[0054] Among them, n p is the number of non-starting and ending POI nodes included in the route. is the heat score of the i-th non-starting and ending POI node; the heat score of the non-starting and ending POI node is calculated as shown in formula (3):

[0055]

[0056] Among them, the POI level is defined according to the category of POI nodes for non-starting and ending POI nodes: the level of first-class POI nodes accounts for 100%, the level of second-class POI nodes accounts for 70%, and the level of third-class POI nodes accounts for 50%;

[0057] The POI level is defined for non-starting and ending POI nodes according to the reputation of the POI nodes: the level of A-level POI nodes accounts for 100%, the level of B-level POI nodes accounts for 70%, and the level of C-level POI nodes accounts for 50%;

[0058] S road is the score of the width suitability of the track route, with a full score of 100, as shown in formula (4):

[0059]

[0060] Among them, n r is the number of paths contained in the track route, S roadi is the width suitability score of the ith path; the width suitability score of the ith path is defined as shown in formula (5):

[0061]

[0062] A path consists of one or more segments. ≥12 is the total length of the road sections with a width greater than or equal to 12 meters in the i-th path, with a proportion of 100%, Di ≥9&<12 is the total length of the road sections with a width greater than or equal to 9 meters and less than 12 meters in the i-th path, accounting for 70%, Di <9 is the total length of the road sections with a width less than 9 meters in the i-th path, accounting for 50% of the total length. sum is the total length of all sections of the i-th path;

[0063] S cross is the smoothness index score of the track route, with a full score of 100, as shown in formula (6):

[0064]

[0065] Where n c is the number of paths contained in the track route, S crossi is the smooth index score of the i-th path; the path smooth index score is defined as shown in formula (7):

[0066]

[0067] The path consists of one or more road segments, DCi4 is the total length of the road segments with a congestion level of "severe congestion" in the i-th path, accounting for 10%, DCi3 is the total length of the road segments with a congestion level of "congestion" in the i-th path, accounting for 50%, DCi2 is the total length of the road segments with a congestion level of "slow moving" in the i-th path, accounting for 70%, DCi1 is the total length of the road segments with a congestion level of "smooth" in the i-th path, accounting for 100%, Di sum is the total length of all sections of the i-th path;

[0068] is the POI density score, and the calculation formula is shown in formula (8):

[0069]

[0070] Where n p is the number of non-starting and ending POI nodes in the route;

[0071] R poi is the proportion of the preset POI heat value dimension, R road is the proportion of the preset width suitability dimension, R cross is the proportion of the preset congestion index dimension; R poi +R road +R cross The sum is 1, where R poi >R road +R cross Indicates the heat value priority recommendation strategy; R road >R poi +R cross Indicates the width suitability priority recommendation strategy; R cross >R poi +R road Indicates the priority recommendation strategy when the route is unobstructed.

[0072] For example, the evaluation value is calculated for each track route in the track route list L obtained in step S4. First, route SA1C1B1E is evaluated: if building A is a 985 university and is rated as first-class A, then If Landmark B is a large square, it is classified as Class II B. If Scenic Spot C is a national 5A scenic spot and is rated as first-class B, then If route SA1 has a total length of 12 km, of which the length of the section with a width of more than 12 meters is 2 km, the length of the section with a width of 9 to 12 meters is 6 km, and the length of the section with a width of 8 to 9 meters is 4 km, then The length of the "severely congested" section is 0 km, the length of the "congested" section is 1 km, the length of the "slow-moving" section is 2 km, and the length of the "unimpeded" section is 9 km. The same goes for If the proportions of POI heat value, width suitability, and route accessibility index are set to R poi =50%, R road =30% and R cross = 20%, then the final evaluation score of route SA1C1B1E is:

[0073]

[0074]

[0075]

[0076]

[0077]

[0078] Similarly, the evaluation values ​​of other track routes can be calculated, and the track routes can be sorted from large to small according to the evaluation values ​​to obtain the final track recommended route list.

[0079] The present invention discloses an intelligent recommendation method for a city marathon track route, which solves the problems of low planning efficiency and many safety hazards caused by the current manual track planning. The present invention can ensure that the planned track passes through many scenic spots, city landmarks, colleges and universities and other areas and buildings with city characteristics, enrich the cultural connotation of the track, promote the city's image, save track planning time, reduce the safety hazards caused by track planning, and at the same time, the planned track route avoids traffic congestion sections, reducing the impact on urban traffic and citizens' travel.

[0080] Embodiment 2

[0081] like Figure 4 As shown, the embodiment of the present invention provides an intelligent recommendation system for urban marathon tracks and routes, including the following modules:

[0082] The data collection module 51 is used to establish a city road information database and a city POI database;

[0083] The starting and ending point selection module 52 is used to determine the starting and ending points of the marathon track according to the city POI database and determine the feasibility of route planning;

[0084] The POI node network construction module 53 is used to obtain the POI data of the starting point and its surroundings, and the POI data of the end point and its surroundings, and filter the POI data according to the urban road information database to obtain the POI node network topology relationship list and construct the POI node network;

[0085] The track route list determination module 54 is used to start from the starting point and use a greedy algorithm to select the adjacent node closest to the current node as the next POI node in the POI node network until the end point is reached; a backtracking algorithm is used to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible to form a track route list.

[0086] The track route recommendation module 55 is used to traverse the track route list, calculate the evaluation value of each track route from four dimensions: POI heat value, width suitability, route accessibility index and POI density, and obtain the final track recommendation route list after sorting according to the evaluation value.

[0087] The above embodiments are provided only for the purpose of describing the present invention, and are not intended to limit the scope of the present invention. The scope of the present invention is defined by the appended claims. Various equivalent substitutions and modifications made without departing from the spirit and principles of the present invention should all be included within the scope of the present invention.

Claims

1. A method for intelligently recommending routes for a city marathon track, characterized in that: include: Step S1: Establishing a city road information database and a city POI database; Step S2: determining the starting point and the end point of the marathon track according to the city POI database, and judging the feasibility of route planning; Step S3: acquiring the POI data of the starting point and its surroundings, and the POI data of the end point and its surroundings, obtaining a POI node network topology relationship list, and constructing a POI node network according to the urban road information database; Step S4: starting from the starting point, using a greedy algorithm in the POI node network to select an adjacent node closest to the current node as the next POI node until the end point is reached; The backtracking algorithm is used to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible to form a track route list, specifically including: Using a greedy algorithm starting from the starting point, select a POI node that is not traversed by the current route and is closest to the current node as the next POI node, and record the distances between the POI nodes that have been traversed until the last POI node is the end point, forming a track route record; when the distance of the track route is greater than 42.695 kilometers or less than 41.695 kilometers, use a backtracking algorithm to backtrack each track route record, use a greedy algorithm to select a route from the end point, form a sub-route, and save the fallback state; fall back in sequence according to the fallback state until it falls back to the starting point; select the track routes with a distance of 41.695 to 42.695 kilometers to form a list of the track routes; Step S5: traverse the track route list, calculate the evaluation value of each track route from four dimensions: POI heat value, width suitability, route accessibility index, and POI density, and sort according to the evaluation value to obtain the final track recommended route list, which specifically includes: The evaluation value calculation formula (1) is as follows: (1) in, is the final evaluation score of the said track route; is the score of the POI heat value dimension of the track route, with a full score of 100, as shown in formula (2): (2) in, is the number of non-starting and ending POI nodes included in the route. is the heat score of the i-th non-starting and ending POI node; the heat score of the non-starting and ending POI node is calculated as shown in formula (3): 100 POI node ranking ratio POI node level ratio (3) The POI levels are defined for non-starting and ending POI nodes according to the categories of the POI nodes: the first-class POI nodes account for 100%, the second-class POI nodes account for 70%, and the third-class POI nodes account for 50%; The POI level is defined for non-starting and ending POI nodes according to the reputation of the POI nodes: the level of A-level POI nodes accounts for 100%, the level of B-level POI nodes accounts for 70%, and the level of C-level POI nodes accounts for 50%; is the score of the width suitability of the track route, with a full score of 100, as shown in formula (4): (4) in, is the number of paths contained in the track route, is the width suitability score of the i-th path; the width suitability score of the i-th path is defined as shown in formula (5): (5) A route consists of one or more segments. is the total length of the road sections with a width greater than or equal to 12 meters in the i-th path, with a proportion of 100%. is the total length of the road sections in the i-th path whose width is greater than or equal to 9 meters and less than 12 meters, accounting for 70% of the total length. is the total length of the road sections with a width less than 9 meters in the i-th path, accounting for 50% of the total length. is the total length of all sections of the i-th path; is the smoothness index score of the track route, with a full score of 100, as shown in formula (6): (6) in is the number of paths contained in the track route, is the smooth index score of the i-th path; the path smooth index score is specified as shown in formula (7): (7) A route consists of one or more road segments. is the total length of the road section with "severe congestion" in the i-th path, accounting for 10% of the total length. is the total length of the road sections with congestion level of "congested" in the i-th path, accounting for 50% of the total length. is the total length of the road section with congestion level of "slow moving" in the i-th path, accounting for 70%, is the total length of the road section with a congestion level of "smooth" in the i-th path, with a proportion of 100%. is the total length of all sections of the i-th path; is the POI density score, and the calculation formula is shown in formula (8): (8) in is the number of non-starting and ending POI nodes in the route; is the proportion of the preset POI heat value dimension, is the proportion of the preset width suitability dimension, is the proportion of the preset congestion index dimension; The sum is 1, where Indicates the heat value priority recommendation strategy; Indicates the width suitability priority recommendation strategy; Indicates the priority recommendation strategy when the route is unobstructed.

2. The method for intelligently recommending urban marathon routes according to claim 1, characterized in that: The step S1: establishing a city road information database and a city POI database, specifically includes: Step S11: collecting urban road information data, and collecting urban road traffic congestion data at different frequencies according to different time periods; building an urban road information database; Step S12: obtaining POI data of POI type "square" as the starting point and end point data sets, obtaining POI data within a straight-line distance of 42.195 kilometers as the surrounding POI data set according to the starting point data set and the end point data set, and constructing a city POI database.

3. The method for intelligently recommending urban marathon routes according to claim 1, characterized in that: The step S2: determining the starting point and the end point of the marathon track according to the city POI database, and judging the feasibility of route planning, specifically includes: Select the starting point POI data and the ending point POI data from the city POI database. If the straight-line distance between the two is less than 21 kilometers, go to step S3, otherwise reselect.

4. The method for intelligently recommending urban marathon routes according to claim 1, characterized in that: The step S3: obtaining the POI data of the starting point and its surroundings, and the POI data of the end point and its surroundings, and filtering the POI data according to the urban road information database, obtaining a POI node network topology relationship list, and constructing a POI node network, specifically includes: Step S31: According to the selected starting point POI data and the ending point POI data, the surrounding POI data within 42.195 kilometers corresponding to the two are obtained from the city POI database; Step S32: According to the urban road information database, the surrounding POI data of the road adjacent to the one-way lane with a total width of less than 8 meters is eliminated to obtain cleaned POI data; Step S33: according to the starting point POI data, the end point POI data and the cleaned POI data, the area around each POI node is divided into four areas, and other POI nodes closest to each area of ​​the POI node are found, and connections are established with them to form a topological relationship, and a POI node network topological relationship list is obtained; Step S34: traverse the POI node network topology relationship list in sequence to obtain route information to adjacent nodes; and filter the road POI node network topology relationship list according to the urban road information database to obtain a node path information list and construct a POI node network.

5. An intelligent recommendation system for city marathon routes, characterized in that: Includes the following modules: Data collection module, used to establish urban road information database and urban POI database; A start and end point selection module is used to determine the start and end point of the marathon track according to the city POI database, and to determine the feasibility of route planning; A POI node network construction module is used to obtain the POI data of the starting point and its surroundings, and the POI data of the end point and its surroundings, obtain a POI node network topology relationship list, and construct a POI node network according to the urban road information database; A track route list determination module is used to select, starting from the starting point, a neighboring node closest to the current node in the POI node network using a greedy algorithm as the next POI node until the end point is reached; The backtracking algorithm is used to ensure that the current route can be rolled back to the previous state, thereby generating multiple sub-routes to include as many POI nodes as possible to form a track route list, specifically including: Using a greedy algorithm starting from the starting point, select a POI node that is not traversed by the current route and is closest to the current node as the next POI node, and record the distances between the POI nodes that have been traversed until the last POI node is the end point, forming a track route record; when the distance of the track route is greater than 42.695 kilometers or less than 41.695 kilometers, use a backtracking algorithm to backtrack each track route record, use a greedy algorithm to select a route from the end point, form a sub-route, and save the fallback state; fall back in sequence according to the fallback state until it falls back to the starting point; select the track routes with a distance of 41.695 to 42.695 kilometers to form a list of the track routes; The track route recommendation module is used to traverse the track route list, calculate the evaluation value of each track route from four dimensions: POI popularity value, width suitability, route accessibility index, and POI density, and obtain the final track recommended route list after sorting according to the evaluation value, which specifically includes: The evaluation value calculation formula (1) is as follows: (1) in, is the final evaluation score of the said track route; is the score of the POI heat value dimension of the track route, with a full score of 100, as shown in formula (2): (2) in, is the number of non-starting and ending POI nodes included in the route. is the heat score of the i-th non-starting and ending POI node; the heat score of the non-starting and ending POI node is calculated as shown in formula (3): 100 POI node ranking ratio POI node level ratio (3) The POI levels are defined for non-starting and ending POI nodes according to the categories of the POI nodes: the first-class POI nodes account for 100%, the second-class POI nodes account for 70%, and the third-class POI nodes account for 50%; The POI level is defined for non-starting and ending POI nodes according to the reputation of the POI nodes: the level of A-level POI nodes accounts for 100%, the level of B-level POI nodes accounts for 70%, and the level of C-level POI nodes accounts for 50%; is the score of the width suitability of the track route, with a full score of 100, as shown in formula (4): (4) in, is the number of paths contained in the track route, is the width suitability score of the i-th path; the width suitability score of the i-th path is defined as shown in formula (5): (5) A route consists of one or more segments. is the total length of the road sections with a width greater than or equal to 12 meters in the i-th path, with a proportion of 100%. is the total length of the road sections in the i-th path whose width is greater than or equal to 9 meters and less than 12 meters, accounting for 70% of the total length. is the total length of the road sections with a width less than 9 meters in the i-th path, accounting for 50% of the total length. is the total length of all sections of the i-th path; is the smoothness index score of the track route, with a full score of 100, as shown in formula (6): (6) in is the number of paths contained in the track route, is the smooth index score of the i-th path; the path smooth index score is specified as shown in formula (7): (7) A route consists of one or more road segments. is the total length of the road section with "severe congestion" in the i-th path, accounting for 10% of the total length. is the total length of the road sections with congestion level of "congested" in the i-th path, accounting for 50% of the total length. is the total length of the road section with congestion level of "slow moving" in the i-th path, accounting for 70%, is the total length of the road section with a congestion level of "smooth" in the i-th path, with a proportion of 100%. is the total length of all sections of the i-th path; is the POI density score, and the calculation formula is shown in formula (8): (8) in is the number of non-starting and ending POI nodes in the route; is the proportion of the preset POI heat value dimension, is the proportion of the preset width suitability dimension, is the proportion of the preset congestion index dimension; The sum is 1, where Indicates the heat value priority recommendation strategy; Indicates the width suitability priority recommendation strategy; Indicates the priority recommendation strategy when the route is unobstructed.

Images