Information pushing method and device based on vehicle position clustering, equipment and medium
By grid division and clustering of vehicle historical trajectory points, the target vehicle is determined and relevant information is pushed, the problems of information push lag and singularity in the existing technology are solved, and accurate information push of the vehicle's driving trajectory is realized.
Patent Information
- Application Number
- CN202411497283.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-24
- Publication Date
- 2025-03-07
AI Technical Summary
The existing vehicle information push method mainly relies on information recommendation when the vehicle is in a designated position, and cannot meet the push needs of the diversified vehicle driving trajectory, and there is a lag in information push.
By obtaining the historical trajectory points of multiple vehicles for grid division, grid-intensive units are extracted, and clustering these units using a clustering model based on dense grid training samples to obtain the clustering results of the vehicles. Then, the target vehicle to be pushed is determined based on the clustering results, and the corresponding push information is retrieved for pushing.
Accurate clustering of vehicle locations is realized, the lag of information push is avoided, and the need for precise push to vehicles of different clusters is met, thus realizing accurate information push based on the vehicle's driving trajectory.
Smart Images
Figure CN119577235A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to an information push method, device, equipment and medium based on vehicle position clustering. Background Art
[0002] With the advent of the big data era, all walks of life are gradually using the collected real-time data for targeted processing to meet higher-level business processing needs. For example, vehicle operators collect vehicle location information to predict vehicle driving needs, driver needs, etc., in order to make accurate recommendations on corresponding information.
[0003] At present, when existing vehicles push information, they usually use the information of the surrounding environment configuration after the vehicle is at a specified location to make recommendations. However, since the surrounding environment of the specified location is fixed, the recommended information is relatively single and cannot meet the push requirements of the diversity of vehicle driving trajectories. In addition, there is a lag in recommending information using only the specified location, which cannot meet the needs of information push with vehicle flexibility. Therefore, there is an urgent need for information push based on vehicle driving trajectories to solve the above problems. Summary of the invention
[0004] In view of this, the present invention provides an information push method, device, equipment and medium based on vehicle position clustering, the main purpose of which is to solve the problem of poor accuracy of existing information recommendation based on vehicle driving positions.
[0005] According to one aspect of the present invention, there is provided an information push method based on vehicle location clustering, comprising:
[0006] Acquire historical trajectory points of multiple vehicles, and divide the historical trajectory points into grids to extract grid dense units;
[0007] Performing clustering processing on the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense centers in the dense grid training samples;
[0008] The target vehicle to be pushed is determined, and the push information corresponding to the target vehicle is retrieved using the clustering result for push.
[0009] Furthermore, the gridding of the historical trajectory points to extract dense grid units includes:
[0010] Determine a map scale and a division size of the historical track point, and divide the historical track point into grids according to the map scale and the division size to obtain grid units;
[0011] The grid density of the grid cell is determined based on the number of trajectory points in the grid cell, and when the grid density is greater than a density threshold, the grid cell is determined to be a grid dense cell.
[0012] Furthermore, before clustering the grid dense units based on the clustering model that has completed model training to obtain the clustering result of the vehicle, the method further includes:
[0013] Acquire dense grid training samples, wherein the dense grid training samples include grid dense unit samples marking the grid dense center;
[0014] The cluster center distance is determined based on the grid dense centers of different grid dense unit samples, and the cluster model is trained using the dense gateway training samples, so that the cluster model training is completed when the cluster center distance remains unchanged.
[0015] Further, after determining the grid density of the grid unit based on the number of trajectory points in the grid unit, the method further includes:
[0016] Performing descending processing on the grid density to obtain a descending grid density;
[0017] If the difference between the density of adjacent descending grids is greater than a preset difference threshold, the density threshold is determined by using the ratio of the density of the descending grids to the number of dimensionality reduction sequences.
[0018] Further, the determining of the target vehicle to be pushed includes:
[0019] The target vehicle to be pushed is extracted from the clustering result according to the push object strategy, wherein the push object strategy includes at least one of a time screening strategy, a location screening strategy, a business screening strategy, and a driving behavior screening strategy.
[0020] Further, the using the clustering result to retrieve the push information corresponding to the target vehicle and push it includes:
[0021] Parsing the location information and clustering category in the clustering result, and obtaining the historical push information matching the target vehicle;
[0022] querying the push information matching the location information, the cluster category and the historical push information based on a preset push mapping relationship, wherein the preset push mapping relationship includes push information to be pushed corresponding to different location information, different cluster categories and different historical push information;
[0023] The push information is pushed to the target vehicle.
[0024] Furthermore, the method further comprises:
[0025] Obtaining a response result of the push information and a triggering position point of the response result;
[0026] The information set in the push pool is updated based on the response result, and the cluster center distance corresponding to the dense grid training sample is updated using the trigger position point to re-execute the model training of the cluster model.
[0027] According to another aspect of the present invention, there is provided an information push device based on vehicle location clustering, comprising:
[0028] An acquisition module, used for acquiring historical trajectory points of multiple vehicles, and performing grid division on the historical trajectory points to extract grid dense units;
[0029] A processing module, configured to perform clustering processing on the grid dense units based on a clustering model that has completed model training, to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on a dense grid training sample, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training sample;
[0030] The push module is used to determine the target vehicle to be pushed, and use the clustering result to retrieve the push information corresponding to the target vehicle for push.
[0031] Furthermore, the acquisition module is specifically used to determine the map scale and division size of the historical trajectory points, and to grid the historical trajectory points according to the map scale and the division size to obtain grid units; determine the grid density of the grid unit based on the number of trajectory points in the grid unit, and when the grid density is greater than a density threshold, determine that the grid unit is a grid dense unit.
[0032] Furthermore, the device further comprises: a training module,
[0033] The acquisition module is used to acquire dense grid training samples, wherein the dense grid training samples include grid dense unit samples marking the grid dense center;
[0034] The training module is used to determine the cluster center distance based on the grid dense centers of different grid dense unit samples, and train the cluster model using the dense gateway training samples, so as to complete the cluster model training when the cluster center distance remains unchanged.
[0035] Furthermore, the device also includes:
[0036] The determination module is used to perform descending processing on the grid density to obtain the descending grid density; if the difference between adjacent descending grid densities is greater than a preset difference threshold, the density threshold is determined by using the ratio of the descending grid density to the dimension reduction sequence number.
[0037] Furthermore, the push module is also used to extract the target vehicle to be pushed from the clustering results according to the push object strategy, and the push object strategy includes at least one of a time screening strategy, a location screening strategy, a business screening strategy, and a driving behavior screening strategy.
[0038] Furthermore, the push module is also used to parse the location information and clustering categories in the clustering results, and obtain historical push information matching the target vehicle; query the push information matching the location information, clustering categories and historical push information based on a preset push mapping relationship, wherein the preset push mapping relationship includes push information to be pushed corresponding to different location information, different clustering categories and different historical push information; and push the push information to the target vehicle.
[0039] Furthermore, the device further comprises: an updating module,
[0040] The acquisition module is further used to acquire the response result of the push information and the triggering position point of the response result;
[0041] The updating module is used to update the information set in the push pool based on the response result, and to update the cluster center distance corresponding to the dense grid training sample using the trigger position point, so as to re-execute the model training of the clustering model.
[0042] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein at least one executable instruction is stored in the computer-readable storage medium, and the executable instruction enables a processor to perform operations corresponding to the above-mentioned information push method based on vehicle location clustering.
[0043] According to another aspect of the present invention, there is provided a computer device, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;
[0044] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned information push method based on vehicle location clustering.
[0045] By means of the above technical solution, the technical solution provided by the embodiment of the present invention has at least the following advantages:
[0046] The present invention provides a method, device, equipment and medium for information push based on vehicle position clustering. Compared with the prior art, the embodiments of the present invention obtain historical trajectory points of multiple vehicles, and grid-divide the historical trajectory points to extract grid dense units; cluster the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the clustering center distance of the clustering model is determined based on the grid dense center in the dense grid training samples; determine a target vehicle to be pushed, and use the clustering result to retrieve push information corresponding to the target vehicle and push it, thereby achieving precise optimization of clustering analysis through gridding, and effectively clustering vehicles according to their positions, avoiding the lag of vehicle information push, meeting the demand for precise push to vehicles in different clusters, and thus achieving the purpose of precise information push based on vehicle driving trajectories.
[0047] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:
[0049] Figure 1 A flow chart of an information push method based on vehicle location clustering provided by an embodiment of the present invention is shown;
[0050] Figure 2 A block diagram showing the composition of an information push device based on vehicle location clustering provided by an embodiment of the present invention is shown;
[0051] Figure 3 A schematic structural diagram of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION
[0052] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.
[0053] The embodiments of the present invention can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.
[0054] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.
[0055] Based on this, in one embodiment, an embodiment of the present invention provides an information push method based on vehicle location clustering, which is illustrated by taking the application of the method to a server and other computer devices as an example, wherein the server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), as well as big data and artificial intelligence platforms, such as advertising push platforms.
[0056] The embodiment of the present invention provides an information push method based on vehicle location clustering, such as Figure 1 As shown, the method includes:
[0057] 101. Obtain historical trajectory points of multiple vehicles, and divide the historical trajectory points into grids to extract grid dense units.
[0058] In an embodiment of the present invention, the vehicle can be a vehicle operating on urban roads or a non-operating vehicle. As the current execution end, it can be an information push server for operating vehicles or a data management server for non-operating vehicles, etc., so as to obtain the historical driving data of the vehicle in real time, and then use data analysis methods to determine the information to be pushed to vehicle users. Among them, the historical trajectory points are multiple trajectory points formed by collecting the vehicle position within a certain historical time period, and are identified by coordinate position points, and can be collected based on the Global Positioning System or the Beidou positioning system. The embodiment of the present invention does not make specific limitations. Since the historical trajectory points are identified based on coordinate position points, in order to cluster the trajectory points, the current execution end first divides the historical trajectory points into grids to obtain multiple grid units containing trajectory points. At this time, the grid dense unit is the unit with a larger number of trajectory points in the grid unit, and the number of trajectory points can be judged based on a predefined density threshold. The embodiment of the present invention does not make specific limitations.
[0059] 102. Perform clustering processing on the grid dense unit based on the clustering model that has completed model training to obtain the clustering result of the vehicle.
[0060] In an embodiment of the present invention, the clustering model can be the K-MEANS algorithm, the K-MEDOIDS algorithm, or the CLARANS algorithm. Preferably, it is the K-MEANS algorithm. The K-MEANS algorithm is applied to the geographic information system to perform clustering analysis on a large amount of geospatial data and identify the characteristics of different regions. Specifically, the K-MEANS algorithm constructs K groups by setting a data set with N tuples, and each group represents a cluster, where K < N. These K groups satisfy the following conditions: (1) Each group contains at least one data record; (2) Each data record belongs to and only belongs to one group. At this time, for the constructed K, the K-MEANS algorithm uses the initial grouping method to perform the first clustering, and then changes the grouping through repeated iteration, so that the grouping scheme after each improvement is better than the previous one, thereby completing the clustering grouping. In an embodiment of the present invention, the clustering model adopted is trained based on dense grid training samples. At this time, the clustering center distance of the clustering model is determined based on the grid dense center in the dense grid training samples, so as to improve the non-uniformity of clustering the trajectory points.
[0061] 103. Determine the target vehicle to be pushed, and use the clustering result to retrieve the push information corresponding to the target vehicle and perform the push.
[0062] In the embodiment of the present invention, after the current execution end obtains the clustering result, the clustering result may include track points that are clustered into multiple categories. The current execution end may determine the corresponding target vehicle for the information to be pushed based on the track points. At this time, the target vehicles corresponding to different clusters may retrieve different information to be pushed for accurate push. For example, if the clustering result is a popular pick-up point, the passenger flow peak time list corresponding to the popular pick-up point may be retrieved and pushed to the vehicle driver, which is not specifically limited in the embodiment of the present invention.
[0063] It should be noted that since the current execution end is the main body for executing steps 101-103 in the embodiment of the present invention, correspondingly, when the current execution end pushes information, it can push it to the client used by the vehicle user, or it can push it to the terminal device installed on the vehicle so that the vehicle driver can view it in real time. The embodiment of the present invention does not make specific limitations.
[0064] The embodiment of the present invention provides an information push method based on vehicle position clustering. Compared with the prior art, the embodiment of the present invention obtains historical trajectory points of multiple vehicles, and grids the historical trajectory points to extract grid dense units; clusters the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training samples; determines a target vehicle to be pushed, and uses the clustering result to retrieve push information corresponding to the target vehicle and push it, thereby achieving precise optimization of clustering analysis through gridding, and effectively clustering vehicles according to their positions, avoiding the lag of vehicle information push, and meeting the demand for precise push to vehicles in different clusters, thereby achieving the purpose of precise information push based on vehicle driving trajectories.
[0065] In another embodiment of the present invention, for further definition and explanation, the step of gridding the historical trajectory points to extract the grid dense units includes:
[0066] Determine a map scale and a division size of the historical track point, and divide the historical track point into grids according to the map scale and the division size to obtain grid units;
[0067] The grid density of the grid cell is determined based on the number of trajectory points in the grid cell, and when the grid density is greater than a density threshold, the grid cell is determined to be a dense grid cell.
[0068] In order to improve clustering efficiency, enable it to process large-scale data, and eliminate local dependencies, the current execution end, when performing grid division, specifically, first determines the map scale of the historical trajectory points. Among them, since the trajectory points of the vehicle are located and collected according to the position points of the vehicle on the map during driving, the historical trajectory points are represented by map data. For example, the position coordinates collected by the vehicle at different timestamps are used as a trajectory point. At this time, the marked position coordinates can be determined as the map scale of the trajectory point, such as uniformly selecting one-hundredth of the real map size as the map scale to perform grid division for the trajectory points in the image. The embodiment of the present invention does not make specific restrictions. In addition, the division size is the length and width of the grid for gridding the trajectory points for map comparison. It can be configured based on different map scales, so as to flexibly segment the data to improve the clustering effect. The embodiment of the present invention does not make specific restrictions. Furthermore, the historical trajectory points marked in the image are grid-divided according to the map scale and grid size to obtain multiple grid units. After obtaining multiple grid units, the grid density of the grid unit g is determined based on the number of trajectory points in the grid unit, that is, the number of trajectory points divided into one grid unit is calculated. If the grid unit g i Density is expressed in Den(g i ) means that it is equal to the grid unit g i The number of data points contained in is compared with the density threshold. When the grid density is greater than the density threshold, it means that the grid contains highly dense data points. Therefore, such grid cells are determined as dense grid cells. That is, if the grid cell density Den(g i ) is greater than the set density threshold Minpts, then the grid cell belongs to the dense cell e, and the dense cell set E = {e1, ..., e n}. If the grid cell is not a dense grid cell, the data points in the grid are free data.
[0069] It should be noted that the division of grid units does not only involve flat division of all historical trajectory points in a location area, but also includes overlapping division, that is, the multiple grid units divided contain trajectory points of overlapping parts, so as to perform density calculation as multiple grid units. At this time, the more grid units are divided, the better the clustering effect, but the model processing pressure is greater. Therefore, the number of grid units can be configured based on clustering requirements, and the embodiment of the present invention does not make specific limitations.
[0070] In another embodiment of the present invention, for further definition and explanation, before the step of clustering the grid dense units based on the clustering model for which model training has been completed to obtain the clustering result of the vehicle, the method further includes:
[0071] Get dense grid training samples;
[0072] The cluster center distance is determined based on the grid dense centers of different grid dense unit samples, and the cluster model is trained using the dense gateway training samples, so that the cluster model training is completed when the cluster center distance remains unchanged.
[0073] In order to use the clustering model to perform cluster analysis on the grid units, thereby improving the clustering efficiency, and achieving the purpose of reducing the number of iterations and quickly finding the cluster center point, the current execution end first obtains dense grid training samples. Among them, the dense grid training samples include grid dense unit samples with marked grid dense centers, which are used as learning basis to train the clustering model. At this time, the clustering model is the K-MEANS algorithm. At this time, if the dense grid unit e i Contains data set X i ={x i1 , x i2 ,…xin},n is equal to Den(ei), then the center of the dense grid cell The set of all dense grid cell centers b is B.
[0074] It should be noted that the clustering model is trained using the dense gateway training sample so that the clustering model training is completed when the cluster center distance remains unchanged. The steps include: 1. Gridding the data in the dense grid training sample; 2. Classifying the data objects of the dense grid training sample into corresponding grid cells; 3. Calculating the density of each grid cell, marking the grids with grid cell density greater than Minpts as dense grid cells, and marking the data in the cells with density less than Minpts as "free data"; 4. Randomly selecting the center of a dense cell from the dense cells as the initial cluster center point c1; 5. Calculating the shortest distance between the center of each dense cell and the currently marked cluster center, that is, the distance to the nearest cluster center, represented by d(b, C): Then calculate the probability that the center of gravity of each dense cell is selected as the next cluster center: 6. Repeat step 5 until K cluster centers are selected; 7. For each free data and dense grid unit center, calculate its distance to the K cluster centers and divide it into the cluster with the smallest distance; 8. For each category C i , suppose there are m free points {x i1 , x i2 , …x im}, and n dense units {e i1 ,…,e in}, the corresponding unit is density {b i1 , ..., b in}Recalculate its cluster center: 9. Repeat steps 7 and 8 until the position of the cluster center no longer changes.
[0075] In another embodiment of the present invention, for further definition and explanation, after the step of determining the grid density of the grid unit based on the number of trajectory points in the grid unit, the method further includes:
[0076] Performing descending processing on the grid density to obtain a descending grid density;
[0077] If the difference between the density of adjacent descending grids is greater than a preset difference threshold, the density threshold is determined by using the ratio of the density of the descending grids to the number of dimensionality reduction sequences.
[0078] In order to improve the classification performance of the clustering model and reduce the accuracy of classification using the clustering model, the current execution end performs descending processing on the grid density before using the clustering model for processing to obtain the descending grid density. Among them, the descending processing of the gateway density is to arrange them in order from large to small to obtain the descending grid density. Then, the difference of each adjacent grid density is compared in turn, that is, the density of all grid units Den(g i ) in descending order, the result is RD(g i ), if RD(g M ) and RD(g M+1 ) is greater than the preset difference threshold, indicating that the density of each unit has obvious jumps. Therefore, the density threshold is determined by the ratio of the grid density after descending to the number of dimensionality reduction sequences, that is, Among them, M is the number of dimensionality reduction sequences.
[0079] In another embodiment of the present invention, for further definition and explanation, the step of determining the target vehicle to be pushed includes:
[0080] The target vehicle to be pushed is extracted from the clustering result according to the push object strategy.
[0081] Since the trajectory points include the positions of all vehicles, in order to meet the accuracy of information push of different vehicles, the current execution end determines the target vehicle to be pushed after determining the clustering results. Among them, when the current execution end extracts the target vehicle, it is screened from the clustering results according to the push object strategy. At this time, the clustering results may include but are not limited to stop gathering points (such as popular passenger pick-up points, popular drop-off points), passing road sections (such as popular routes), etc., and the embodiment of the present invention does not make specific restrictions. Among them, the push object strategy includes at least one of the time screening strategy, the location screening strategy, the business screening strategy, and the driving behavior screening strategy. The time screening strategy is a method for screening vehicles according to time conditions, including but not limited to holidays, road section peaks, etc. The location screening strategy is a method for screening vehicles according to designated locations, including but not limited to schools, stations, etc. The business screening strategy is a method for screening vehicles according to specific marketing businesses and specific public welfare businesses, including but not limited to new house sales, public welfare fund activities, etc. The driving behavior screening strategy is a method for screening vehicles according to specific driving behaviors, including but not limited to stop behaviors, waiting behaviors, etc., and the embodiment of the present invention does not make specific restrictions.
[0082] It should be noted that when extracting the target vehicle according to the above-mentioned push object strategy, the time, location, business and driving behavior corresponding to each trajectory point are first analyzed from the clustering results. Specifically, if the clustering results of grid units a, c, and d are stay gathering points, the vehicle corresponding to the trajectory point in this grid unit a, c, and d can be determined as the target vehicle according to the location screening strategy. The embodiment of the present invention does not make specific limitations.
[0083] In another embodiment of the present invention, for further definition and explanation, the step of using the clustering result to retrieve the push information corresponding to the target vehicle and push includes:
[0084] Parsing the location information and clustering category in the clustering result, and obtaining the historical push information matching the target vehicle;
[0085] Querying the push information matching the location information, the cluster category and the historical push information based on the preset push mapping relationship;
[0086] The push information is pushed to the target vehicle.
[0087] In order to meet the flexible push requirements of different information, when the current execution end uses the clustering results to call the push information, it first parses the location information and clustering categories in the clustering results. Among them, the location information is the location of the trajectory point corresponding to the clustering of different types, and the clustering category is the various categories that can be obtained after clustering. The samples can be labeled and learned during model training, and the embodiment of the present invention does not make specific limitations. Furthermore, the current execution end obtains the historical push information that matches the target vehicle, and queries the push information that matches the location information, clustering category and historical push information based on the preset push mapping relationship. At this time, the preset push mapping relationship includes the push information to be pushed corresponding to different location information, different clustering categories and different historical push information. For example, the location information is a railway station site, the clustering category is a frequent boarding point, and the historical push information is a city map and scenic spot information, which is pushed to the target vehicle for playback, and the embodiment of the present invention does not make specific limitations.
[0088] In another embodiment of the present invention, for further definition and description, the steps further include:
[0089] Obtaining a response result of the push information and a triggering position point of the response result;
[0090] The information set in the push pool is updated based on the response result, and the cluster center distance corresponding to the dense grid training sample is updated using the trigger position point to re-execute the model training of the cluster model.
[0091] In order to make the clustering model have better learning ability and improve the clustering effect, after pushing the information, the current execution end obtains the response result for the pushed information, such as viewing the behavior as the response result and feeding it back to the current execution end. At the same time, the trigger location point that triggers this response result is obtained. The location information can be collected by the global positioning system while receiving the response result to obtain the trigger location point. The embodiment of the present invention does not make specific limitations. Furthermore, the information set in the push pool is updated based on the response result, and the cluster center distance corresponding to the dense grid training sample is updated using the trigger location point to re-execute the model training of the clustering model. Among them, when the cluster center distance is updated using the trigger location point, that is, the dense grid unit center b' is updated according to the trigger location point and added to the set B to re-execute the model training of the clustering model.
[0092] The embodiment of the present invention provides an information push method based on vehicle position clustering. Compared with the prior art, the embodiment of the present invention obtains historical trajectory points of multiple vehicles, and grids the historical trajectory points to extract grid dense units; clusters the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training samples; determines a target vehicle to be pushed, and uses the clustering result to retrieve push information corresponding to the target vehicle and push it, thereby achieving precise optimization of clustering analysis through gridding, and effectively clustering vehicles according to their positions, avoiding the lag of vehicle information push, and meeting the demand for precise push to vehicles in different clusters, thereby achieving the purpose of precise information push based on vehicle driving trajectories.
[0093] Furthermore, as a response to the above Figure 1 The implementation of the method shown in the figure, the embodiment of the present invention provides an information push device based on vehicle location clustering, such as Figure 2 As shown, the device comprises:
[0094] An acquisition module 21 is used to acquire historical track points of multiple vehicles, and to perform grid division on the historical track points to extract grid dense units;
[0095] A processing module 22 is used to perform clustering processing on the grid dense units based on the clustering model that has completed model training to obtain the clustering result of the vehicle, wherein the clustering model is obtained by training based on the dense grid training sample, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training sample;
[0096] The push module 23 is used to determine the target vehicle to be pushed, and use the clustering result to retrieve the push information corresponding to the target vehicle for push.
[0097] Furthermore, the acquisition module is specifically used to determine the map scale and division size of the historical trajectory points, and to grid the historical trajectory points according to the map scale and the division size to obtain grid units; determine the grid density of the grid unit based on the number of trajectory points in the grid unit, and when the grid density is greater than a density threshold, determine that the grid unit is a grid dense unit.
[0098] Furthermore, the device further comprises: a training module,
[0099] The acquisition module is used to acquire dense grid training samples, wherein the dense grid training samples include grid dense unit samples marking the grid dense center;
[0100] The training module is used to determine the cluster center distance based on the grid dense centers of different grid dense unit samples, and train the cluster model using the dense gateway training samples, so as to complete the cluster model training when the cluster center distance remains unchanged.
[0101] Furthermore, the device also includes:
[0102] The determination module is used to perform descending processing on the grid density to obtain the descending grid density; if the difference between adjacent descending grid densities is greater than a preset difference threshold, the density threshold is determined by using the ratio of the descending grid density to the dimension reduction sequence number.
[0103] Furthermore, the push module is also used to extract the target vehicle to be pushed from the clustering results according to the push object strategy, and the push object strategy includes at least one of a time screening strategy, a location screening strategy, a business screening strategy, and a driving behavior screening strategy.
[0104] Furthermore, the push module is also used to parse the location information and clustering categories in the clustering results, and obtain historical push information matching the target vehicle; query the push information matching the location information, clustering categories and historical push information based on a preset push mapping relationship, wherein the preset push mapping relationship includes push information to be pushed corresponding to different location information, different clustering categories and different historical push information; and push the push information to the target vehicle.
[0105] Furthermore, the device further comprises: an updating module,
[0106] The acquisition module is further used to acquire the response result of the push information and the triggering position point of the response result;
[0107] The updating module is used to update the information set in the push pool based on the response result, and to update the cluster center distance corresponding to the dense grid training sample using the trigger position point, so as to re-execute the model training of the clustering model.
[0108] The embodiment of the present invention provides an information push device based on vehicle position clustering. Compared with the prior art, the embodiment of the present invention obtains historical trajectory points of multiple vehicles, and grids the historical trajectory points to extract grid dense units; clusters the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training samples; determines a target vehicle to be pushed, and uses the clustering result to retrieve push information corresponding to the target vehicle and push it, thereby achieving precise optimization of clustering analysis through gridding, and effectively clustering vehicles according to their positions, avoiding the lag of vehicle information push, and meeting the demand for precise push to vehicles in different clusters, thereby achieving the purpose of precise information push based on vehicle driving trajectories.
[0109] According to one embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores at least one executable instruction, and the computer-executable instruction can execute the information push method based on vehicle position clustering in any of the above method embodiments.
[0110] Figure 3 A schematic diagram of the structure of a computer device provided according to an embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computer device.
[0111] like Figure 3 As shown, the computer device may include: a processor 302 , a communication interface 304 , a memory 306 , and a communication bus 308 .
[0112] The processor 302 , the communication interface 304 , and the memory 306 communicate with each other via the communication bus 308 .
[0113] The communication interface 304 is used to communicate with other devices such as clients or other servers.
[0114] The processor 302 is used to execute the program 310, and specifically can execute the relevant steps in the above-mentioned embodiment of the information push method based on vehicle location clustering.
[0115] Specifically, the program 310 may include program codes, which include computer operation instructions.
[0116] The processor 302 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiment of the present invention. The one or more processors included in the computer device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.
[0117] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.
[0118] The program 310 may be specifically configured to enable the processor 302 to perform the following operations:
[0119] Acquire historical trajectory points of multiple vehicles, and divide the historical trajectory points into grids to extract grid dense units;
[0120] Performing clustering processing on the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense centers in the dense grid training samples;
[0121] The target vehicle to be pushed is determined, and the push information corresponding to the target vehicle is retrieved using the clustering result for push.
[0122] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.
[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A method for pushing information based on vehicle location clustering, characterized in that: include: Acquire historical trajectory points of multiple vehicles, and divide the historical trajectory points into grids to extract grid dense units; Performing clustering processing on the grid dense units based on a clustering model that has completed model training to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on dense grid training samples, and the cluster center distance of the clustering model is determined based on the grid dense centers in the dense grid training samples; The target vehicle to be pushed is determined, and the push information corresponding to the target vehicle is retrieved using the clustering result for push.
2. The method according to claim 1, characterized in that The gridding of the historical trajectory points to extract dense grid units includes: Determine a map scale and a division size of the historical track point, and divide the historical track point into grids according to the map scale and the division size to obtain grid units; The grid density of the grid cell is determined based on the number of trajectory points in the grid cell, and when the grid density is greater than a density threshold, the grid cell is determined to be a dense grid cell.
3. The method according to claim 1, characterized in that Before clustering the grid dense units based on the clustering model that has completed model training to obtain the clustering result of the vehicle, the method further includes: Acquire dense grid training samples, wherein the dense grid training samples include grid dense unit samples marking the grid dense center; The cluster center distance is determined based on the grid dense centers of different grid dense unit samples, and the cluster model is trained using the dense gateway training samples, so that the cluster model training is completed when the cluster center distance remains unchanged.
4. The method according to claim 2 or 3, characterized in that: After determining the grid density of the grid unit based on the number of trajectory points in the grid unit, the method further includes: The grid densities are processed in descending order to obtain descending grid densities; If the difference between the density of adjacent descending grids is greater than a preset difference threshold, the density threshold is determined by using the ratio of the density of the descending grids to the number of dimensionality reduction sequences.
5. The method according to claim 1, characterized in that Determining the target vehicle to be pushed includes: The target vehicle to be pushed is extracted from the clustering result according to the push object strategy, wherein the push object strategy includes at least one of a time screening strategy, a location screening strategy, a business screening strategy, and a driving behavior screening strategy.
6. The method according to claim 5, characterized in that The using the clustering result to retrieve the push information corresponding to the target vehicle and push it includes: Parsing the location information and clustering category in the clustering result, and obtaining the historical push information matching the target vehicle; querying the push information matching the location information, the cluster category and the historical push information based on a preset push mapping relationship, wherein the preset push mapping relationship includes push information to be pushed corresponding to different location information, different cluster categories and different historical push information; The push information is pushed to the target vehicle.
7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: Obtaining a response result of the push information and a triggering position point of the response result; The information set in the push pool is updated based on the response result, and the cluster center distance corresponding to the dense grid training sample is updated using the trigger position point to re-execute the model training of the cluster model.
8. An information push device based on vehicle location clustering, characterized in that: include: An acquisition module, used for acquiring historical trajectory points of multiple vehicles, and performing grid division on the historical trajectory points to extract grid dense units; A processing module, configured to perform clustering processing on the grid dense units based on a clustering model that has completed model training, to obtain a clustering result of the vehicle, wherein the clustering model is obtained by training based on a dense grid training sample, and the cluster center distance of the clustering model is determined based on the grid dense center in the dense grid training sample; The push module is used to determine the target vehicle to be pushed, and use the clustering result to retrieve the push information corresponding to the target vehicle for push.
9. A computer-readable storage medium, wherein at least one executable instruction is stored in the computer-readable storage medium, and the executable instruction enables a processor to execute operations corresponding to the information push method based on vehicle location clustering as described in any one of claims 1 to 7.
10. A computer device comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the information push method based on vehicle location clustering as described in any one of claims 1-7.